Skip to main content

Full text of "CMH Pub 5-5 Cartwheel: The Reduction Of Rabaul"

See other formats


The War in the Pacific 



This volume, one of the series UNITED STATES ARMY IN 
WORLD WAR II, is the eighth to be published in the subseries 
THE WAR IN THE PACIFIC. All the volumes will be closely 
related, and the series will present a comprehensive account of the 
activities of the Military Establishment during World War II. A 
tentative list of subseries is appended at the end of this volume. 

Library of Congress Catalog Card Number: 59-60004 

For tale by the Superintendent of Document!, U.S. Government Printing Office 
Waihingion, D.C. 20402 


Kent Roberts Greenfield, General Editor 

Elmer Ellis 
University of Missouri 

Samuel Flagg Bemis 
Yale University 

Gordon A. Craig 
Princeton University 

Oron J. Hale 
University of Virginia 

W. Stull Holt 
University of 

Advisory Committee 

(Ai of jo May 1958) 

Brig. Gen. John B. Sullivan 


Brig. Gen. Edgar C. Dolcinan 
Army War College 

Brig. Gen, Frederick R. Zierath 
Command and General Staff College 

Industrial College of the Armed Forces 

Col. Vincent J. Esposito 
United States Military Academy 

T. Harry Williams 
Louisiana State University 

Office of the Chief of Military History 
Maj. Gen. Richard W. Stephens, Chief 

Chief Historian Kent Roberts Greenfield 

Chief, Histories Division Col. Seneca W. Foote 

Chief, Editorial and Publication Division Lt, Col. E. E. Steck 

Editor in Chief Joseph R. Friedman 

Chief, Cartographic Branch Elliot Dunay 

Chief, Photographic Branch Margaret E. Tackky 

The History of 


prepared under the direction of Louis Morton 

The Fall of the Philippines 
Guadalcanal: The First Offensive 
Victory in Papua 
Cartwheel: The Reduction of Rabaul 
Seizure of the Gilberts and Marshalls 
Campaign in the Marianas 
The Approach to the Philippines 
Leyte: The Return to the Philippines 
Triumph in the Philippines 
Okinawa: The Last Battle 
Strategy and Command: Turning the Tide, 1941-1943 
Strategy and Command: The Road to Victory, 1943-1945 

. . . to Those Who Served 


The campaign described in the present volume was important to the 
Army as an experience in amphibious warfare and combined operations 
against a formidable and still resourceful enemy. It was also of critical 
importance in the evolution of American strategy in the Pacific. Cartwheel 
began as an uphill fight with means that seemed inadequate to the ends 
proposed, even though these were limited. But it swiftly brought our forces 
to a crest from which we were able to launch the two powerful drives, 
through the Southwest and Central Pacific, that crushed Japan before we 
redeployed the forces directed against Germany. The campaign put to the 
test the principle of unity of command, and also the capacity for co-opera- 
tion between two theaters, one under Army, the other under Navy com- 
mand, and both under forceful and dominant commanders. By ingenious 
and aggressive use of the ground, sea, and air forces at their disposal they 
made these suffice to achieve more than had been foreseen as possible, and 
opened up a new vista of strategy. They took a heavy toll of the enemy's 
resources, established the technique of bypassing his strongholds, includ- 
ing finally Rabaul itself, and threw him on the defensive. This book will 
be of interest not only to professional officers, but also to a wide variety of 
other readers and students. 

Washington, D. C. R. W. STEPHENS 

30 May 1958 Maj. Gen., U. S. A. 

Chief of Military History 


The Author 

Born in Scotland and a U.S. citizen since 1928, John Miller, jr., was 
awarded the degree of Doctor of Philosophy in History by the State Uni- 
versity of Iowa in 1942. In World War II he saw service overseas with the 
U.S. Marine Corps in New Zealand and in the Solomon Islands, where he 
participated in the Bougainville operations described in this volume. A 
member of the historical staff of the Department of the Army since 1945, 
Dr. Miller is the author of Guadalcanal; The First Offensive in the present 
series, coauthor of Korea: 1951-1953, and contributor of several chapters 
to the 1956 edition of ROTC Manual 145-20, American Military History, 
160J-1953. He has written articles and reviews for historical and military 
journals, and has taught history at the University of Omaha, the State 
University of Iowa, the Graduate School of the U.S. Department of Agri- 
culture, and The American University in Washington, D.C. 



The reduction of Rabaul was accomplished by a gigantic double en- 
velopment which required closely co-ordinated land, sea, and air opera- 
tions by the armed forces of the United States and her Pacific allies. This 
volume, like the others in the series, attempts to explain in detail the part 
played by the U.S. Army ground forces and to make clear, by summary, 
the contributions of all forces and nations. 

The Cartwheel battles differed from those of the two earlier cam- 
paigns, Guadalcanal and Papua, that were directed toward the reduction 
of Rabaul. In Guadalcanal and Papua the antagonists, more evenly matched 
than in later campaigns, strained themselves to bring relatively small 
ground forces to bear on narrow fronts, so that great issues hinged on the 
outcome of regimental and battalion actions. A study of those campaigns, 
therefore, quite properly focuses on tactics. During the period covered by 
this book the Allied commanders could employ superior forces over a vast 
area while the Japanese had no recourse but to entrench themselves in an 
effort to hold out and inflict as many casualties as possible. This volume 
attempts to analyze the techniques by which the Allies employed their 
strength to bypass fortified positions and seize weakly defended but stra- 
tegically important areas, or, in the apt baseball parlance used by General 
MacArthur, to "hit 'em where they ain't." It is, therefore, a study in strategy 
and high command as well as in tactics. 

The willing, able counsel and assistance I have received in preparing 
this book have greatly eased my task. Dr. Louis Morton, Chief of the Pacific 
Section of the Office of Military History during the period of research and 
writing, and my other friends and colleagues in this Office have aided 
unstintingly. Dr. Kent Roberts Greenfield, Chief Historian of the Army, 
has been a constant source of wise and kindly help. The successive Chiefs 
of Military History— Maj. Gens. Orlando Ward, Albert C. Smith, John H. 
Stokes, and Richard W. Stephens— and Cols. Thomas J. Sands, George G. 
O'Connor, Ridgway P. Smith, Jr., and Seneca W. Foote have appreciated 
the nature and worth of history and provided encouragement and powerful 

For locating and furnishing to me, without restriction, all the neces- 
sary records I wish to make public my gratitude to the efficient records 


staff of this Office and of the Military Records Branch, Federal Records 
Center, of the U.S. General Services Administration; the Historical Branch, 
G— 3, Headquarters, U.S. Marine Corps; and the Naval History Division 
of the Office of the Chief of Naval Operations. I also owe thanks to Messrs. 
Stanley L. Falk and Thomas G. Wilds for performing research and trans- 
lation in Japanese records, to Mrs. Marguerite Bartz for typing the manu- 
script, and to the participants named on pp. 386-87 who generously read all 
or parts of the manuscript and sent in helpful comments and additional 

Final editing was the responsibility of Mrs. Gay Morenus Hammerman, 
who also prepared the index. Mrs. Nancy Easterling Payne was copy editor. 
Maps were prepared under the supervision of Maj. James F. Holly and 
Mr. Elliot Dunay. Miss Margaret E. Tackley selected the photographs and 
wrote the captions. To these capable and friendly colleagues who con- 
tributed so much— many thanks. 

Responsibility for any deficiencies in this book is mine alone. 

Washington, D. C. JOHN MILLER, jr. 

30 May 1958 



Chapter Page 


Early Pacific Strategy i 

The Casablanca Conference 6 


Preliminary Theater Planning 9 

The Pacific Military Conference 11 

Preparation of the Directive of 28 March 1943 15 


The Southwest Pacific Area 20 

The Plan of Maneuver . 25 


Japanese Command and Strategy 32 

Japanese Offensives, January-June 1943 36 

Japanese Strength and Dispositions, 30 June 1943 45 



Nassau Bay 59 


South Pacific Organization 67 

Preparations and Plans 70 

Secondary Landings 78 

Rendova 85 

The Move to Zanana 92 

Rice Anchorage . , . . 94 


Japanese Plans 97 

Operations of the Northern Landing Group 99 

Operations of the Southern Landing Group 106 

Casualties 120 

Command and Reinforcements 122 


Chapter Page 


The Attack on Bairoko 127 

Pressure on the Japanese 132 

Preparations for the Corps Offensive 137 


Plans 143 

Ilangana and Shimizu Hill: The 43d Division 146 

The Attack Against the Ridges: The 3Jth Division .... 149 

Capture of the Airfield 158 


The Airfield 165 

Reinforcements 167 

The Cleanup 169 

Vella Lavella: The Bypass 171 

Final Operations 184 


Plans 189 

Allied Air and Naval Preparations 195 

Lae: The Seaborne Invasion 202 

Nadzab: The Airborne Invasion 207 

Strategic Reconsiderations 212 

Advance Through the Ramu Valley 216 

The Coastal Advance 217 


The Decision To Bypass Rabaul 222 

The General Plan 225 

Air Operations in October 229 

Forces and Tactical Plans 234 

Preliminary Landings 239 

Seizure of Empress Augusta Bay 241 


Air and Surface Action, 1-11 November 251 

Operations Ashore 255 

December Attacks Against Rabaul 269 


Plans and Preparations 272 

Arawe 282 

Cape Gloucester 289 

Saidor 295 


Chapter Page 


General Plans 306 

Reducing Rabaul and Kavieng 309 

Seizure of the Green Islands 312 


The Decision 316 

The Reconnaissance in Force 321 

To the Shores of Seeadler Harbour 332 

Lorengau 339 


Preparations 351 

Hill 700 358 

Hill 260 364 

Action by the Creeks 371 





INDEX 399 


No. Page 

1. Comparison of Allied Intelligence Estimates With Japanese Strength 

and Dispositions, Southeast Area, 30 June 1943 47 

2. American Casualties on New Georgia 187 


1. Organization of Forces for Cartwheel 21 

2. Estimated Timing and Sequence of Cartwheel Operations .... 28 

3. Organization of Japanese Forces, Southeast Area, June 1943 ... 33 

4. Southwest Pacific Organization for Woodlark-Kiriwina 52 

5. Organization of Principal South Pacific Forces, June 1943 .... 68 


No. Page 

6. Organization of South Pacific Forces for Toenails 74 

7. Organization of Attack Force, D Day 77 

8. Western Force on D Day 79 

9. Eastern Force on D Day 80 

10. South Pacific Organization for Vella Lavella Invasion 176 

11. Organization of Northern Force [TF 31], Vella Lavella 177 


a. The Pacific Areas, as of 1 August 1942 3 

3. The Cartwheel Area 23 

4. The Wau Area 37 

5. Operation Chronicle Area, 30 June 1943 50 

6. Southern Approaches to Salamaua 62 

7. Landings in New Georgia, 21 June-5 July 1943 69 

8. Approach to Bairoko, 5-20 July 1943 100 

9. Drive Towards Munda Point, 2-14 July 1943 107 

10. Capture of Munda Point, 22 July-4 August 1943 145 

11. The Cleanup, 5-27 August 1943 166 

12. The Huon Peninsula and the Straits 189 

13. Opening the Markham Valley, 4-16 September 1943 ...... 204 

14. Capture of Finschhafen, 22 September-20 October 1943 217 

15. Bougainville Landings, 27 October— 1 November 1943 225 

16. Situation on Bougainville, 15 December 1943 258 

17. Arawe Landings, 15 December 1943 283 

18. Cape Gloucester Landings, 26-29 December 1943 292 

19. Seeadler Harbour Area 318 

20. Los Negros Assault, 29 February-g March 1944 324 

21. Lugos Mission to Lorengau, 15-18 March 1944 341 

22. Japanese Counterattack on Bougainville, 9—17 March 1944 .... 353 


General Douglas MacArthur 4 

Some Pacific Planners in Conference 17 

Vice Adm. Jinichi Kusaka 35 

General Hitoshi Imamura 35 

Lt. Gen. Hatazo Adachi 36 



Lt. Gen. Haruyoshi Hyakutake 38 

Japanese Troop Transport Under Attack 40 

Brig. Gen. Nathan F. Twining 51 

Troops Disembarking From LCI . 57 

Natives Carrying Luggage 58 

Jeep and Trailer Leaving an LST gg 

Clearing Airfield Site With Hand Tools 60 

Airfield at Segi Point 8a 

Men of i52d Field Artillery Battalion 84 

Ships Moving Toward Rendova 86 

Aboard the Transport McCawley 87 

Men of 43d Signal Company Wading Ashore 89 

Truck Towing a 155-mm. Howitzer Over Muddy Trail 93 

Maj. Gen. Noboru Sasaki 98 

Troops of the 173d Infantry Wading Across a Creek 111 

Evacuating Casualties, 12 July 1943 113 

Jeep Trail From Zanana 115 

Infantry Loading on LCP (R)'s 116 

LCM's Approaching Laiana, New Georgia 117 

Rear Adm. Theodore S. Wilkinson 125 

Japanese Prisoners Captured Near Laiana Beach 128 

Pillbox Made of Coconut Logs and Coral 133 

Soldiers of the 161st Infantry 138 

Bombing of Munda Airfield, Early Morning 141 

Munda Airfield . 160 

Reducing an Enemy Pillbox With a Flame Thrower 162 

Light Tanks M3 of the 9th Marine Defense Battalion 164 

Munda Airfield in Operation 168 

4-Ton Truck Stuck in the Mud 170 

37th Division Troops Carrying Weapons and Ammunition 171 

Warship Firing at Japanese Destroyers 182 

14th New Zealand Brigade Group 185 

Maj. Gen. J. Lawton Collins 188 

B-24 Over Salamaua 197 

Enemy Aircraft Destroyed on the Ground 198 

B-25 Medium Bombers 199 

Salamaua 202 

Crossing Rain-Swollen Francisco River 203 

Australian Troops Debarking From LST's 206 

C-47 Transport Planes Loaded With Parachute Troops 208 

Airdrop at Nadzab, Morning of 5 September 1943 209 

Bombing Rabaul 231 

B-25's Leaving Bougainville 233 

Lt. Gen. Alexander A. Vandegrift 237 



Maj. Gen. Allen H. Turnage 243 

Mount Bagana 245 

3d Marines Landing on Cape Torokina 247 

LCVP's on the Beach at Empress Augusta Bay 249 

Aircrewman Wounded in Strike on Rabaul 254 

Amphibian Tractors LVT (1) 256 

Tractor and Trailer in Mud 257 

Soldiers of the 148th Regimental Combat Team 259 

Admiral William F. Halsey, Jr 262 

37th Division Troops 263 

Results of Japanese Bombing of Puruata Island 264 

105-mm. Howitzer 266 

4.2-Inch Chemical Mortar 267 

Lt. Gen. Millard F. Harmon 268 

C-47 Air-Dropping Supplies 270 

B-25's Over Wewak 281 

Alligator Returning to Beach on Arawe 289 

Early Morning Bombardment 291 

7th Marines Landing on Narrow Beach 293 

Rear Adm. Daniel E. Barbey 297 

M10 Motor Carriage Mounting 3-Inch Gun 301 

Japanese Ships Burning at Rabaul 310 

Aboard the Cruiser Phoenix 325 

First Wave of Landing Craft Unloading 327 

2d Lt. Marvin J. Henshaw 328 

Digging a Foxhole Through Coral Rock 329 

Lt. Gen. Walter Krueger 337 

LST's Loaded With Troops and Equipment 339 

Men of the 8th Cavalry 346 

Crossing the Lorengau River 348 

Troop G, 8th Cavalry, Near Number 1 Road 349 

60-mm. Mortar Emplacement 354 

155-mm. Guns of the 3d Marine Defense Battalion 357 

37th Division Men Carrying 5-Gallon Cans of Water 360 

Two Light Tanks M3 of the 754th Tank Battalion 363 

"OP Tree" on Hill 260 365 

South Knob, Hill 260 370 

North Knob, Hill 260 371 

Maj. Gen. Robert S. Beightler 376 

Tank-Infantry Attack 377 

Japanese Pillbox on Fire 378 

All illustrations are from Department of Defense files. 



The Strategic Background 

The great Japanese bastion at Rabaul 
on New Britain in the Bismarck Archi- 
pelago posed a double threat to the Allies 
from 1942 through the early months of 
1944. Bristling with warships and air- 
planes, it menaced the line of commu- 
nications from the United States to Aus- 
tralia, and it blocked any Allied advance 
along the north coast of New Guinea to 
the Philippines. Reduction of Rabaul 
was therefore the primary mission, dur- 
ing this period, of the Allied forces of 
the South and Southwest Pacific Areas. 
In executing this mission these forces 
fought a long series of ground, air, and 
naval battles spaced across a vast region. 

Early Pacific Strategy 

Before the Allies could move effec- 
tively against Rabaul itself, they had to 
clear the way by seizing Guadalcanal 
and driving the Japanese out of the Pa- 
puan Peninsula. With the successful con- 
clusion of these two campaigns in early 
1943, the South and Southwest Pacific 
forces completed the first phase of a series 
of offensive operations against Rabaul 
that had been ordered by the U.S. Joint 
Chiefs of Staff in July 1943. The strategic 
purpose of this series was defensive, the 
scale limited. The immediate aim of the 
Joint Chiefs was, not to defeat the Japa- 

nese nation, but to protect Australia and 
New Zealand by halting the Japanese 
southward advance from Rabaul toward 
the air and sea lines of communication 
that joined the United States and 
Hawaii to Australia and New Zealand. 

These orders stemmed from earlier, 
more fundamental decisions by Presi- 
dent Franklin D. Roosevelt, Prime Min- 
ister Winston S. Churchill, and the U.S.- 
British Combined Chiefs of Staff, who 
from the very outset had agreed to de- 
feat Germany first and then to concen- 
trate against Japan. Pending Germany's 
defeat, the Allies decided on a defensive 
attitude in the Pacific. But within this 
framework they firmly resolved that Aus- 
tralia, New Zealand, the Hawaiian Is- 
lands, and Midway were not to be al- 
lowed to fall into Japanese hands. 1 

Throughout the early months of 1943 
the Japanese threat to the Allied line 
of communications had mounted stead- 

1 For complete discussions on the development 
of this strategy see Maurice Matloff and Edwin 
M. Snell, Strategic Planning for Coalition War- 
fare: 1941-1942 (Washington, 1953), Chs. I-VIII; 
Louis Morton, The Fall of the Philippines (Wash- 
ington, 1953), Chs. II-IV, IX; and Mark Skinner 
Watson, Chief of Staff: Prewar Plans and Prepara- 
tions (Washington, 19.150), pp. 367-521. All are in 
WAR II. See also Louis Morton's volumes on 
strategy, command, and logistics in the Pacific, now 
in preparation for the same series. 



ily. The enemy's capture of Rabaul in 
January placed him in an excellent posi- 
tion to move south. Well situated in 
relation to Truk and the Palau Islands, 
Rabaul possessed a magnificent harbor 
as well as sites for several airfields. Only 
440 nautical miles southwest of Rabaul 
lies the New Guinea coast, while Guadal- 
canal is but 565 nautical miles to the 
southeast. Thus the Japanese could ad- 
vance southward covered all the way by 
land-based bombers. And since none of 
the islands in the Bismarck Archipelago- 
New Guinea— Solomons area lay beyond 
fighter-plane range of its neighbors, the 
Japanese could also cover their advance 
with fighters by building airstrips as they 
moved along. By May 1942 they had 
completed the occupation of the Bis- 
marck Archipelago. They pushed south 
to establish bases at Lae and Salamaua 
on the northeast coast of New Guinea, 
and built airfields in the northern Solo- 

With the Japanese seemingly able to 
advance at will, the Joint Chiefs had 
been making all possible efforts to pro- 
tect Hawaii, Midway, New Zealand, and 
Australia by holding the lines of com- 
munication. Troops to reinforce existing 
Allied bases and to establish new bases 
were rushed overseas in early 1942. The 
g2d and 41st Divisions went to Australia. 
The 37 th Division was dispatched to the 
Fijis, the Americal Division to New 
Caledonia, and the 147th Infantry to 
Tongatabu. Troops of the Americal Di- 
vision, plus Navy and Marine units, oc- 
cupied posts in the New Hebrides begin- 
ning in March. A Navy and Marine force 
held Samoa. 

At this time the Japanese planned to 
cut the line of communications and iso- 

late Australia by seizing the Fijis, Samoa, 
New Caledonia, and Port Moresby in 
New Guinea. But even before they were 
turned back from Port Moresby by the 
Allies during May, in the naval battle of 
the Coral Sea, the Japanese had post- 
poned the attacks against the Fijis, New 
Caledonia, and Samoa and had planned 
instead the June attempt against Mid- 
way. Although they managed to seize a 
foothold in the Aleutians, they failed 
disastrously at Midway. With four air- 
craft carriers sunk and hundreds of 
planes and pilots lost, the Japanese could 
no longer continue their offensives. The 
Allies were thus able to take the initia- 
tive in the Pacific. 

To conduct operations, the Joint 
Chiefs organized the Pacific theater along 
lines which prevailed for the rest of the 
period of active hostilities. By agreement 
in March 1942 among the Allied nations 
concerned, they set up two huge com- 
mands, the Southwest Pa cific Area and 


the Pacific Ocean Area. 2 (Map 
Southwest Pacific included Australia and 
adjacent waters, all the Netherlands 
Indies except Sumatra, and the Philip- 
pine Islands. 

The vast Pacific Ocean Areas em- 
braced nearly all the remainder of the 
Pacific Ocean. Unlike the Southwest 
Pacific, which was one unit, the Pacific 
Ocean Areas were divided into three 
parts— the South, Central, and North Pa- 
cific Areas. The North Pacific included 
the ocean reaches north of latitude 42 

s The plural is customarily employed for the 
Pacific Ocean Areas, although the JCS directive 
establishing the command used "Area." See CCS 
57/1, Memo, JCS for President, 30 Mar 42, title: 
Dirs to CINCPOA and to the Supreme Comdr 
SWPA, with Incls. 



# Johnston* 

MAP 2 

north; the Central Pacific lay between 
42 north and the Equator. 

The South Pacific Area, which lay 
south of the Equator, east of longitude 
159 east, and west of longitude no° 
west, was an enormous stretch of water 
and islands that included but one mod- 
ern sovereign nation, the Dominion of 
New Zealand. Among the islands, many 
of them well known to readers of ro- 
mantic fiction, were the French colony 
of New Caledonia, the British-French 

Condominium of the New Hebrides, 
and the Santa Cruz, Fiji, Samoa, Tonga, 
Cook, Society, and Marquesas Islands. 
The boundary separating the South and 
Southwest Pacific Areas (longitude 159 
east) split the Solomon Islands. 

General Douglas MacArthur, Supreme 
Commander or, as he came to be called, 
Commander in Chief of the Southwest 
Pacific Area, with headquarters at Bris- 
bane, Australia, in early 1943, com- 
manded all land, air, and sea forces 



General Douglas MacArthur, Commander in Chief of the Southwest Pacific 
Area, with Admiral Chester W. Nimitz, Commander in Chief of the Pacific 
Ocean Areas. Photograph taken in Brisbane, Australia, March 1944. 

assigned by the several Allied govern- 
ments. 3 This famous and controversial 
general was enjoined from directly com- 
manding any national force. In contrast 
Admiral Chester W. Nimitz, who was 
concurrently Commander in Chief of 
the Pacific Ocean Areas, with authority 
over all Allied forces assigned, was also 
Commander in Chief of the Pacific Fleet, 
He exercised direct control over the 
North and Central Pacific Areas but in 
accordance with the Joint Chiefs' in- 

3 "Supreme Commander" was the title used by 
CCS 57/1, 30 Mar 4a. MacArthur seems to have 
preferred "Commander in Chief" and "Supreme 
Commander" fell into disuse. 

structions appointed a subordinate as 
commander of the South Pacific Area 
with headquarters first at Auckland, 
New Zealand, and later at Noumea, New 
Caledonia. Like MacArthur, this officer 
was ineligible to command any national 
force directly. Admiral William F. Hal- 
sey, Jr., the incumbent at the close of 
the Guadalcanal Campaign, replaced the 
original commander, Vice Adm. Robert 
L. Ghormley, on 18 October while the 
campaign was reaching its climax. 

At the time of the Coral Sea engage- 
ment in May, a small Japanese force had 
garrisoned Tulagi in the Solomons, and 
shortly afterward the Japanese began 



building an airfield at nearby Lunga 
Point on Guadalcanal. Just before they 
learned of the Japanese airfield under 
construction on Guadalcanal, the Joint 
Chiefs capitalized on the Midway vic- 
tory by ordering the South and South- 
west Pacific Areas to begin the advance 
against Rabaul. The operations, as set 
forth in the Joint Chiefs' orders of 2 
July 1942, were divided into three 
phases. The first, or "Task One," was 
the seizure of Tulagi and Guadalcanal 
in the Solomons, and of the Santa Cruz 
Islands. Since possession of the Santa 
Cruz Islands did not prove necessary, 
they were never taken. Task Two in- 
cluded the capture of the remainder of 
the Japanese-held Solomons and of Lae, 
Salamaua, and other points on the north- 
east coast of New Guinea in the South- 
west Pacific Area. Task Three was the 
seizure and occupation of Rabaul itself, 
and of adjacent positions. 4 

Command during Task One, which 
would be executed in the South Pacific 
Area, was entrusted to the South Pacific 
commander. Tasks Two and Three, to 
be carried out by South and Southwest 
Pacific Area forces entirely within the 
Southwest Pacific Area, were to be con- 
ducted under MacArthur's command. 

When they received the Joint Chiefs' 
directive, the commanders of the South 
and Southwest Pacific Areas met in Mel- 
bourne, Australia, to discuss the three 
tasks. They agreed that the advance 
should be governed by two basic con- 
cepts: the progressive forward movement 

"Jt Dir for Offen Opns in SWPA Agreed on by 
U.S. CsofS, 2 Jul 42, OPD 381, Sec 2, Case 83. 
Unlike other JCS directives, this paper bore no 
JCS number. It is also reproduced in JCS 112, 
21 Sep 42, title: Mil Sit in the Pac. 

of air forces and the isolation of Rabaul 
before the final assault. After the initial 
lunge into Guadalcanal, there would fol- 
low a series of advances £o seize air and 
naval bases in New Guinea, New Britain, 
and the northern Solomons. With these 
bases Allied fighter planes and bombers 
would be in position to cover the en- 
tire Bismarck Archipelago-eastern New 
Guinea-Solomons area and isolate Ra- 
baul from the east, west, north, and south 
before troops were put ashore to capture 
the great base. 5 

The Joint Chiefs of Staff assigned the 
reinforced ist Marine Division as the 
landing force for Task One. That unit, 
landing on Guadalcanal and Tulagi on 
7 August 1942, quickly captured its 
major objectives. The Japanese reaction 
to the invasion was so violent and reso- 
lute, and Allied control over the air and 
sea routes so tenuous, that the campaign 
did not end then but dragged on for six 
months. It was not until February 1943 
—after two Army divisions and one more 
Marine division had been committed to 
the battle and six major naval engage- 
ments fought— that Guadalcanal was 
completely wrested from the enemy. 8 

5 Dispatch, CINCSWPA and COMSOPAC to CofS 
82s, ABC 370.26 (7-8-42), Sec 1. 

6 For the history of the Guadalcanal Campaign 
see John Miller, jr., Guadalcanal: The First Offen- 
II (Washington, ig4g); Samuel Eliot Morison, His- 
tory of United States Naval Operations in World 
War II, Vol. IV, Coral Sea, Midway, and Submarine 
Actions (Boston: Little, Brown and Company, 
1949), and Vol. V, The Struggle for Guadalcanal 
(Boston: Little, Brown and Company, 1949); Maj. 
John L. Zimmerman, USMCR, The Guadalcanal 
Campaign (Washington, 1949); and Wesley Frank 
Craven and James Lea Cate, eds., The Army Air 
Forces in World War II, Vol. IV, The Pacific- 
Guadalcanal to Saipan— August 1942 to July 1944 
(Chicago: The University of Chicago Press, 1950). 



With the Guadalcanal victory, the Allies 
seized the initiative from the Japanese 
and halted their southward advance. The 
Japanese never attempted the assaults 
against the Fijis, Samoa, and New Cale- 

Just as the Guadalcanal Campaign was 
opening, a Japanese force landed at 
Buna, on the northeast coast of New 
Guinea's Papuan peninsula, and at- 
tempted to capture the vital Allied base 
at Port Moresby by crossing the tower- 
ing Owen Stanley Range. But the offen- 
sive stalled, and MacArthur was able to 
move the 32d U.S. Division, the 7th 
Australian Division, and several addi- 
tional American regimental combat 
teams and Australian infantry brigades 
against the Japanese beachheads at Buna, 
Gona, and Sanananda on the Papuan 
peninsula, as well as to establish bases at 
Milne Bay at Papua's tip and on Good- 
enough Island in the D'Entrecasteaux 
group. 7 

At the beginning of 1943, with both 
the Guadalcanal and Papuan campaigns 
drawing to a successful close, the Allies 
could look forward to using Guadalcanal 
and Papua as bases for continuing the 
advance against Rabaul. In the Central 
Pacific, Admiral Nimitz could not under- 
take any offensive westward from Pearl 
Harbor and Midway until the line of 
communications to Australia was abso- 
lutely secure. At this time both Halsey 
and MacArthur were preparing plans for 
their campaigns against Rabaul, but had 
not yet submitted them to the Joint 
Chiefs of Staff. 

7 For a detailed discussion of the war in the 
Southwest Pacific, 1942-43, see Samuel Milner, Vic- 
WORLD WAR II (Washington, 1957). 

The Casablanca Conference 

Although the Joint Chiefs of Staff had 
not yet received detailed plans for Ra- 
baul, they were well aware of the impor- 
tance of the operations in the South 
and Southwest Pacific Areas. These op- 
erations naturally had to be considered 
in the light of global strategy and re- 
viewed by the U.S.-British Combined 
Chiefs of Staff. 8 

By the end of 1942, the Joint Chiefs 
were concluding their study of Allied 
objectives for the year 1943. President 
Roosevelt and the service chiefs were 
then preparing to meet at Casablanca 
in French Morocco with Prime Minister 
Churchill and the British Chiefs in order 
to explore the problem fully and deter- 
mine Allied objectives for the year. No 
final plan for the defeat of Japan had 
been prepared but the subject was being 
studied in Washington. 9 Also under dis- 
cussion were the question of advancing 
against Japan through the North Pacific 

8 See Miller, Guadalcanal: The First Offensive, 
pp. 172-73; Min, JCS mtg, 22 Dec 42; Min, JPS 
nug, 16 Sep 42; JCS 1 12/1, 14 Oct 42; title: Mil Sit 
in the Pac. For a more detailed discussion see Ray 
S. Cline, Washington Command Post: The Opera- 
tions Division, ^UNITED STATES ARMY IN 
WORLD WAR II (Washington, 1951), pp. 215-19, 
and Robert E. Sherwood, Roosevelt and Hopkins: 
An Intimate History (New York: Harper & Brothers, 
1948), Ch. XXVII. See also John Miller, jr., "The 
Casablanca Conference and Pacific strategy," Mili- 
tary Affairs, XIII (Winter 1949), 4. 

Unless otherwise indicated, this section is based 
on the proceedings and papers of the Casablanca 
Conference which are filed in regular sequence 
with the CCS and JCS minutes and papers. They 
were also printed and bound, along with the pro- 
ceedings of the meetings attended by the President 
and Prime Minister, in a useful separate volume- 
Casablanca Conference: Papers and Minutes of 

"JPS 67/2, 4 Jan 43, title: Proposed Dir for a 
Campaign Plan for the Defeat of Japan. 



and the possibility of conducting opera- 
tions in Burma to reopen the road to 
China. 10 

Pacific operations, and the emphasis 
and support that the advance on Rabaul 
would receive, were significantly affected 
by decisions made at Casablanca. During 
the ten-day conference that began on 14 
January the President, the Prime Min- 
ister, and the Combined Chiefs of Staff 
carefully weighed their strategic ends, 
apportioned the limited means available 
to accomplish them, and so determined 
Allied courses of action for 1943. 

The Americans and British who met 
at Casablanca agreed on general objec- 
tives, but their plans differed in several 
important respects. The Americans 
wished the Allies to conduct a strategic 
offensive directly against Germany and 
to aid the Soviet Union, but they also 
favored strong action in the Pacific and 
Far East. It was imperative, in their view, 
to guarantee the security of Allied lines 
of communication there and to break the 
enemy hold on positions that threatened 
them. Convinced that China had to be 
kept in the war, they recommended that 
the British, with the aid of American 
ships and landing craft, recapture Burma 
so that the Burma Road could be re- 
opened and the Allies could send more 
supplies to bolster Generalissimo Chiang 
Kai-shek's armies. They wished to keep 
the initiative in the Southwest and South 
Pacific, to inflict heavy losses on the Japa- 
nese, and eventually to use Rabaul and 
nearby positions as bases for further 
advances. Admiral Ernest J. King, Chief 
of Naval Operations, expressed the hope 

10 Min, JPS mtgs, 2 and g Dec 42; Min, JCS 
mtgs, 25 Aug, 15 Sep, and 15 Dec 42, and 5 Jan 43; 
Min, CCS mtg, 6 Nov 42. 

that 30 percent of Allied military power 
could be deployed against the Japanese 
instead of the 15 percent which he esti- 
mated was then being used. 

The British understandably shied 
away from enlarging the scope of Allied 
action in the Pacific. With the Germans 
right across the Channel from England, 
the British stressed the importance of 
concentrating against Germany first. 
While admitting the necessity for retak- 
ing Burma, they strongly emphasized the 
importance of aiding the Soviet Union. 
They promised to deploy their entire 
strength against Japan after the defeat 
of Germany, and suggested that the Jap- 
anese should meanwhile be contained by 
limited offensives. At the same time the 
British desired to extend the scope of 
Allied operations in the Mediterranean. 

General George C. Marshall, Chief of 
Staff, and Admiral King opposed what 
Marshall called "interminable opera- 
tions in the Mediterranean." They ad- 
vocated maintaining constant, unremit- 
ting pressure against the Japanese to pre- 
vent them from digging in and consoli- 
dating their gains. Warning that the 
American people would not stand for an- 
other Bataan, Marshall argued that suf- 
ficient resources must be kept in the 
Pacific; otherwise "a situation might 
arise in the Pacific at any time that would 
necessitate the United States regretfully 
withdrawing from the commitments in 
the European Theatre." 11 Admiral King, 
pointing out the strategic importance of 
an advance across the Central Pacific to 
the Philippines, raised the question of 
where to go after Rabaul was captured. 
The British did not wish to make spe- 

11 Min, CCS mtg, 17 Jan 43. 



cific commitments for operations beyond 
Rabaul but suggested a meeting after its 
capture to decide the question. 

By 23 January Americans and British 
had reconciled their differences over stra- 
tegic objectives for 1943. They agreed to 
secure the sea communications in the 
Atlantic, to move supplies to the Soviet 
Union, to take Sicily, to continue their 
build-up of forces in Britain for the in- 
vasion of northern France, and— a de- 
cision that was to have a marked effect 
on Pacific operations— to bomb Germany 
heavily in the Combined Bomber Offen- 
sive that was to be launched by mid- 
summer 1943. To make sure that none 
of these undertakings would be jeop- 
ardized by the need for diverting strength 
to prevent disaster in the Pacific, ade- 

quate forces would be maintained in the 
Pacific and Far East. What was consid- 
ered "adequate" was not defined. 

The Combined Chiefs agreed in prin- 
ciple that Burma was to be recaptured 
by the British and that they would meet 
later in the year to make final decisions. 
In the Pacific the Allies were to main- 
tain constant pressure on Japan with the 
purpose of retaining the initiative and 
getting into position for a full-scale of- 
fensive once Germany had surrendered. 
Specifically, the Allies intended to cap- 
ture Rabaul, make secure the Aleutians, 
and advance west through the Gilberts 
and Marshalls in the Central Pacific 
toward Truk and the Marianas. The 
Central Pacific advances were supposed 
to follow the capture of Rabaul. 


Selecting Objectives 

With Allied strategic objectives for 
1943 determined at Casablanca, the next 
task facing the Army and Navy com- 
manders in Washington and in the Pa- 
cific was the selection of exact tactical 
objectives. Two considerations would be 
paramount in making a choice: the mili- 
tary value of the objectives, and the re- 
sources that would be available. The 
process of selection was not completed 
until a full-dress conference involving 
the Joint Chiefs of Staff, the Joint Staff 
Planners, and representatives of all the 
Pacific commands had met in Washing- 
ton. 1 Such a conference was made neces- 
sary by the large disparity between the 
size of the forces General MacArthur 
asked for to take his objectives and the 
size of the forces that were actually avail- 

Preliminary T heater Planning 
General MacArthur' s Plans 

In the Southwest Pacific, General Mac- 
Arthur had begun planning for the of- 
fensive against Rabaul at an early date. 
His plans for Tasks Two and Three— 
mutually supporting advances along two 
axes, culminating in a converging attack 
against Rabaul— were the same in early 
January 1943 as those outlined in July 

J See below, pp. 1 1 l— 1 

of the previous year. But his forces could 
not start Task Two until the Allies had 
successfully completed the Guadalcanal 
and Papuan campaigns in the first two 
months of 1943. There were not enough 
ground troops to undertake any offensive 
moves immediately, MacArthur re- 
ported, and there were far from enough 
air forces to conduct the campaigns. 2 

In order to advance against Rabaul in 
one continuous movement, MacArthur 
wished to assemble all the necessary 
forces before starting the offensive, and 
substantial reinforcements would be re- 
quired. In both South and Southwest 
Pacific Areas there were troops equiva- 
lent to fifteen and two-thirds American, 
New Zealand, and Australian divisions, 
but not all were trained and equipped 
for offensive action. Of the six trained 
Southwest Pacific divisions, five would be 
resting and reorganizing for some time 
to come, after fighting in Guadalcanal, 
Papua, and the Middle East. There were 
seven trained divisions— six American 
and one New Zealand— as well as some 
separate infantry and cavalry regiments 
in the South Pacific. Three of the di- 
visions and one regiment had seen serv- 
ice on Guadalcanal and were enjoying 

! Rad to MacArthur, 7 Jan 43, CM-OUT 2273; 
Rad to MacArthur, 8 Jan 43, CM-OUT 2833; Rad 
From MacArthur, 10 Jan 43, CM-IN 4574. 



a well-deserved rest. The equivalent 
therefore of only five divisions plus sev- 
eral separate regiments could be counted 
as ready for immediate use. 

In naval strength, MacArthur was lim- 
ited to cruisers, destroyers, and subma- 
rines. He had no carriers, no battleships, 
and few cargo ships, transports, and 
landing craft. The greater part of the 
Pacific Fleet, including aircraft carriers, 
battleships, cruisers, and destroyers, was 
operating in the South Pacific Area. 

Both areas boasted a total air strength 
of about 1,850 land-based planes of all 
types— bombers, fighters, and cargo 
planes. These planes came from the U.S. 
Army Air Forces, the U.S. Navy, the 
U.S. Marine Corps, the Royal New Zea- 
land Air Force, and the Royal Aus- 
tralian Air Force. 3 

At this time General MacArthur was 
looking forward to targets well beyond 
Rabaul; he had set his sights on the 
Philippine Islands. In February 1943 
he and his staff concluded that the com- 
pletion of the campaign against Rabaul 
could secure for the Allies "important, 
but not decisive advantages." These ad- 
vantages would certainly aid future oper- 
ations but, except for the destruction of 
precious shipping, would do little dam- 
age to Japan's main economic structure. 
Since the Netherlands Indies contained 
the great economic wealth, especially oil, 
taken by Japan in 1941 and 1942, a 

1 Information on strength and plans at this time 
is drawn from GHQ SWPA. Elkton Plan for the 
Seizure and Occupation of the New Britain-New 
Guinea-New Ireland Area, 12 and 28 Feb 43, and 
from U.S. Strategic Bombing Survey, The Employ- 
ment of Forces Under the Southwest Pacific Com- 
mand (Washington, 1947), p. 18. The latter is an 
almost verbatim copy of a series of monographs pre- 
pared during and immediately after the war by the 
Historical Section, G-3, GHQ, SWPA. 

decisive blow could be struck, Mac- 
Arthur reasoned, by cutting the lines of 
communication between Japan and the 
Indies. As the Philippine Islands lay 
squarely athwart all sea and air routes 
between Japan and the Indies, the Allies 
could cut them by establishing air and 
naval bases in the Philippines. General 
MacArthur concluded that he should 
move to the Philippines by advancing 
westward along New Guinea's north 
coast, then swinging northwest through 
the intermediate islands into the Philip- 
pines. The advance along the New 
Guinea coast might be started about the 
time that the siege of Rabaul began, but 
could not safely start until Rabaul was 
neutralized lest ships and planes based 
there harry or obstruct the advance. 4 

This plan for advancing to the Philip- 
pines, called Reno, had not yet been 
transmitted to Washington. It looked far 
into the future. There were not enough 
forces to inaugurate the Rabaul plan, 
Elkton. Certainly not enough were 
available to begin Reno. 

Admiral Halsey's Plan 

In the South Pacific, Admiral Halsey 
looked on Munda Point in New Georgia 
as the most likely first objective for his 
forces under Task Two. 6 The Japanese 
had started an airfield at the Australian 
Methodist Mission on Munda in Novem- 
ber 1942 when their attempts to recap- 
ture Henderson Field on Guadalcanal 
had faltered. The new field was intended 
to serve as an advanced air base in an- 

* GHQ SWPA, Estimate of the Situation and 
Rough Draft, Reno Plan, 25 Feb 43, OCMH. 

8 The name "Munda" is apparently a phonetic 
rendition of a native term rather than a reflection 
of Caesar's glory. 



other attempt to retake Henderson Field 
in 1943. 

The Japanese exhibited skill and cun- 
ning in concealing their activities at 
Munda. Even though the Allies had long 
known that Munda Point was being used 
as a staging area, they were not sure that 
an airfield was under construction until 
3 December. The Japanese had rigged 
cables to the tops of the palm trees, then 
cut the trunks away and left the cables 
holding up the treetops. Thus hidden 
from aerial observation, they built their 
runway and then cut down the camou- 
flage. The day they completed the run- 
way, 15 December 1942, the Japanese 
decided to build a second airfield at Vila 
on nearby Kolombangara. 6 The airfields 
at Munda and Vila, only 180 nautical 
miles from Henderson Field, presented 
a serious threat to the Allied positions in 
the Solomons and New Hebrides. 

In Allied hands, Munda would be 
invaluable for continuing the advance 
against Rabaul, and Admiral Halsey's 
forces pressed on. They capped their 
success on Guadalcanal with the blood- 
less seizure of the Russell Islands on 21 
February 1943. This shortened the air- 
line distance to Munda by sixty-five 
miles and provided torpedo boat and 
landing craft bases to assist in the capture 
of all New Georgia, an operation then 
being planned by Halsey and his sub- 

Munda Point was, physically, one of 
the best sites for an air base in the Solo- 
mon Islands. Strategically, it was well 

8 Southeast Area Naval Operations, I, Japanese 
Monogr No. 48 (OCMH) in the series Japanese 
Studies in World War II, 47, 52; ONI USN, Combat 
Narratives: Solomon Islands Campaign, X, Opera- 
tions in the New Georgia Area, 21 June— 5 August 
•943 [Washington, 1944], 1-2. 

situated to support an advance to Bou- 
gainville, which would be necessary if 
South Pacific aircraft were to strike 
Rabaul effectively. 7 South Pacific forces 
would have to use aircraft carriers to 
advance directly from Henderson Field 
to Bougainville, but possession of Munda 
Point would enable them to advance 
progressively under cover of land-based 
fighter planes and bombers and obviate 
the need to use precious carriers close 
to islands that were studded with enemy 
airfields. 8 

The Pacific Military Conference 

Neither Mac Arthur nor Halsey could 
start his offensive yet. They had not yet 
agreed on a co-ordinated plan, and they 
lacked enough forces to begin. Allot- 
ment of forces would depend on deci- 
sions by the Joint and Combined Chiefs 
of Staff, who at Casablanca had decided 
on the program for 1943 without know- 
ing how many troops, planes, and ships 
would be needed for Rabaul. 

Shortly before leaving Washington for 
Casablanca, the Joint Chiefs had in- 
structed MacArthur to submit detailed 
plans for carrying out their directive of 
2 July 1942 and authorized him to ex- 
change views with Nimitz and Halsey. 
They suggested personal meetings by the 

' From Henderson Field to the Shortland Islands 
is 285 nautical miles, to Kahili, 300 miles, to Buka 
Passage, 363 miles. From Munda, which is within 
fighter range of Henderson and the Russells, to the 
Shortlands is 100 nautical miles, to Kahili, 125 
miles, to Buka Passage, 234 miles, and to Rabaul, 
394 miles. 

8 Adm William F. Halsey, Jr., Narrative Account 
of the South Pacific Campaign, 3 Sep 44, OCMH; 
Lt Gen Millard F. Harmon, The Army in the South 
Pacific, 6 Jun 44, p. 7, OCMH; Fleet Admiral Wil- 
liam F. Halsey and Lt Comdr J. Bryan, III, Ad- 
miral Halsey's Story (New York: Whittlesey House, 
1947). P- '54- 



commanders or by their staffs to prepare 
a broad plan that would enable the Joint 
Chiefs to give careful consideration to 
such matters as timing, reinforcement, 
supply, and the transfer of command 
over Tasks Two and Three to Mac- 
Arthur. 9 Maintaining that it was incon- 
venient for high commanders to under- 
take long journeys away from their head- 
quarters, MacArthur radioed his ideas 
for Tasks Two and Three to Nimitz and 
Halsey. On 11 February Halsey sent his 
deputy commander, Rear Adm. Theo- 
dore S. Wilkinson, to Brisbane to begin 
a co-ordinated plan. 10 

Shortly thereafter MacArthur asked 
the Joint Chiefs for permission to send 
his chief of staff and several other officers 
to Washington to explain his plans. The 
Joint Chiefs approved, but stipulated 
that representatives from Halsey 's and 
Nimitz' areas should also come for a gen- 
eral discussion of Pacific problems. 11 The 

"Rad to MacArthur, 8 Jan 43, CM-OUT 2833; 
Rad to Maj Gen Rush B. Lincoln, New Caledonia 
(to be passed to Halsey), 8 Jan 43, CM— OUT 2834; 
Rad to Lt Gen Delos C. Emmons, Hawaii (to be 
passed to Nimitz), 8 Jan 43, CM-OUT 2835; Rad 
to MacArthur, 1 1 Jan 43, CM-OUT 3664. 

10 Rad from MacArthur, 27 Jan 43, CM-IN 12553; 
Rad from MacArthur, 11 Feb 43, CM-IN 5610; 
Commander, South Pacific Area and South Pacific 
Force, War Diary: 1 January 1943—30 June 1944 
(hereafter cited as COMSOPAC War Diary), 11- 
12, 14—15 Feb 43 entries. 

"Rad from MacArthur, 15 Feb 43, CM-IN 7418; 
Rads to MacArthur, 16 Feb 43, CM-OUT 5656 and 
CM-OUT 5660; Rads to Harmon (for Halsey), 16 
Feb 43, CM-OUT 5658 and CM-OUT 5661; Rads 
to Emmons (for Nimitz), 16 Feb 43, CM-OUT 5657 
and CM-OUT 5659. The Pacific representatives 
timed their trip to accompany Brig. Gen. Albert C. 
Wedemeyer to Washington. Wedemeyer, a member 
of the Operations Division of the War Department 
General Staff and of the Joint and Combined Staff 
Planners, visited the Southwest Pacific to explain 
the Casablanca decisions to MacArthur and to be- 
come better acquainted with the area. 

delegates reached Washington on 10 
March and two days later met with Ad- 
miral King and various officers from the 
Army and Navy planning and logistical 
staffs. 12 

Thus began the series of meetings, 
generally known as the Pacific Military 
Conference, which were to produce a 
new directive for operations. This con- 
ference constituted an excellent example 
of the detailed and undramatic, but ab- 
solutely essential, spadework that had 
to precede major decisions affecting the 
course of the war in the Pacific. 

The ELKTON Plan 

After Admiral King opened the first 
session on 12 March with a strategic 
review of the world situation, Maj. Gen. 
Richard K. Sutherland, MacArthur's 
chief of staff, presented the Elkton plan 

12 To represent him, Halsey had selected General 
Harmon, Commanding General, U.S. Army Forces 
in the South Pacific Area; Maj. Gen. Nathan F. 
Twining, commanding the, Thirteenth Air Force; 
and two staff officers, Brig. Gen. Dewitt Peck, 
USMC, his war plans officer, and Capt. Miles R. 
Browning, USN, his chief of staff. MacArthur sent 
Maj. Gen. Richard K. Sutherland, his chief of staff; 
Brig. Gen. Stephen J. Chamberlin, his operations 
officer; and Lt. Gen. George C. Kenney, Com- 
mander, Allied Air Forces, SWPA. Representing 
Nimitz were his deputy and chief of staff, Vice Adm. 
Raymond A. Spruance, General Emmons, and Capt. 
Forrest P. Sherman. Present at the first meeting, 
besides King and the Pacific delegates, were Lt. 
Gen. Joseph T. McNarney, Deputy Chief of Staff 
of the Army; Lt. Gen. Stanley D. Embick, of the 
Joint Strategic Survey Committee; Maj. Gen. George 
E. Stratemeyer, Chief of the Air Staff; Maj. Gen. 
Thomas T. Handy, Assistant Chief of Staff, Opera- 
tions Division, War Department General Staff; Maj. 
Gens. LeRoy Lutes and Lucius D. Clay, of Head- 
quarters, Army Service Forces; Vice Adm. Russell 
Willson, of the Joint Strategic Survey Committee; 
Rear Adm. Charles M. Cooke, Jr.; and Wedemeyer. 



to the conference. 13 This plan, bearing 
the date 28 February 1943, was a re- 
vision of the first Elkton plan, which 
was dated 12 February, and prescribed 
the same general scheme of maneuver 
as MacArthur's earlier plans for the re- 
duction of Rabaul. MacArthur had pre- 
pared it on the assumption that he would 
control both the Southwest and South 
Pacific forces for Tasks Two and Three, 
for the Joint Chiefs' directive had stated 
explicitly that these would be conducted 
under his command. Halsey, according 
to MacArthur, had already assented to 

Elkton's intelligence estimate pointed 
out that the Japanese generally con- 
trolled the north coast of New Guinea 
northwest of Buna, as well as New Brit- 
ain, New Ireland, and the Solomons 
northwest of Guadalcanal. Japanese de- 
fenses were concentrated, as were Allied 
holdings in the region, in the vicinity of 
airfields. Except for the perimeters 
around the airfields and naval bases, the 
land areas were pretty well unoccupied. 

Between 79,000 and 94,000 Japanese 
troops were thought to be stationed in 
the New Guinea— Bismarck Archipel- 
ago-Solomons area. Enemy air strength 
was estimated at 383 land-based planes, 
while 4 battleships, 2 aircraft carriers, 
14 cruisers, 11 seaplane tenders, about 
40 destroyers, numerous auxiliaries, and 
about 50 merchant ships of 3,000 tons or 
over were on hand for operations. It 
was expected that the Japanese, if at- 
tacked, could be immediately reinforced 

"Notes on Pac Conf Held in Mar 43, 1st mtg, 12 
Mar 43. George C. Kenney, General Kenney Re- 
ports: A Personal History of the Pacific War (New 
York: Duel), Sloan and Pearce, 1949), devotes one 
chapter (VIII) to this conference. 

by 10,000 to 12,000 troops and about 250 
planes as well as major portions of the 
Combined Fleet from the Netherlands 
Indies, Japanese home waters, and the 
Philippines Islands. In six months, 615 
more aircraft could be committed, and 
10 or 15 divisions might be dispatched if 
shipping was available. 

Having described the forbidding na- 
ture of the enemy stronghold, General 
Sutherland proceeded, in his presenta- 
tion of the Elkton plan, to outline the 
contemplated Allied moves. The execu- 
tion of Tasks Two and Three would re- 
quire mutually supporting, co-ordinated 
advances along two lines: one, by South- 
west Pacific forces in the west, from New 
Guinea to New Britain; the other, by 
South Pacific forces in the east, through 
the Solomons. Elkton broke Tasks Two 
and Three into five operations: 

1. Seizure of airdromes on the Huon 
Peninsula of New Guinea to provide air 
support for operations against New 

2. Seizure of Munda Point as well as 
other airdromes on New Georgia to 
cover operations against New Ireland 
in the Bismarck Archipelago and the 
remainder of the Solomons; 

3. Seizure of airdromes on New Brit- 
ain and Bougainville to support opera- 
tions against Rabaul and Kavieng in 
New Ireland; 

4. Capture of Kavieng and the isola- 
tion of Rabaul, although it was consid- 
ered possible that Kavieng might be 
taken after Rabaul; 

5. Capture of Rabaul. 

The timing of these missions was not 
rigidly fixed, nor was there an estimate 
of the time required to carry them out. 



Large forces, assembled in advance, 
were required to execute the five opera- 
tions of Elkton— and there was the rub 
from the point of view of the Washing- 
ton planners faced with global responsi- 
bilities. They listened as Sutherland read 
a detailed accounting of forces on hand 
and forces requested. The plan, in brief, 
called for five additional divisions, forty- 
five additional air groups, or about twice 
the 1,850 land-based planes then on 
hand, and an unspecified number of war- 
ships, transports, cargo ships, and land- 
ing craft sufficient to mount and support 
all the operations. 14 

The official records do not disclose 
with what emotions the officers from the 
various Washington agencies received 
the information about the necessary re- 
inforcements, but it is not difficult to 
imagine that some were surprised. At 
Casablanca the Americans had assumed 
the capture of Rabaul in 1943 as a mat- 
ter of course, and had confidently dis- 
cussed the possibility of advancing be- 
yond Rabaul. 

The Pacific delegates learned imme- 
diately that there was virtually no chance 
for them to get all the reinforcements 
that they wanted. 15 It was possible to 
effect some increases in the number of 
aircraft, but to give General MacArthur 
everything he asked would have cut too 
deeply into the bomber offensive against 
Germany. There were several trained 
divisions available in the United States, 
but there were not enough transports 
to ship them overseas in time, or to sup- 
ply them after their arrival. Everyone at 

" GHQ SWl'A, Elkton Plan .... 28 Feb 43. 
" Rad, Sutherland to MacArthur, 12 Mar 43, 
CM-OUT 1930. 

the conference was convinced of the 
necessity for offensive operations, but it 
was recognized that the operations would 
be limited by the available means. Ad- 
miral Halsey's representatives, Lt. Gen. 
Millard F. Harmon of the Army, Brig. 
Gen. Dewitt Peck of the Marine Corps, 
and Capt. Miles R. Browning of the 
Navy, endorsed the Elkton plan, but 
some of the Navy planners in Washing- 
ton were dubious of its value. They 
believed it would tie up too many ships 
and too many troops for too long a time, 
and would not achieve decisive results. 
The Washington planners informed the 
Pacific representatives that only two or 
three more divisions and a few more 
planes could be sent overseas. 16 

The solution therefore was to replace 
the ambitious directive of 2 July 1942 
with something more realistic. Before 
deciding on a new directive, the Joint 
Chiefs instructed the Pacific delegates to 
decide what offensive operations they 
thought could be undertaken in 1943 
with the allotted forces. It was under- 
stood that the Pacific commanders would 
not be committed by their subordinates' 
recommendations. 17 

The Pacific delegates answered 
promptly. They stated that the South 
and Southwest Pacific forces would be 
able to advance as far as the southeast 
part of Bougainville, seize eastern New 
Guinea up to Madarig, extend to Wood- 
lark and Kiriwina in the Trobriand 
Islands, and advance to Cape Gloucester 

" For the detailed record of debate and discus- 
sion at the various meetings of the Pacific Military 
Conference, see Notes on Pac Conf Held in Mar 
43, with Inclosures and Annexes. See below, Biblio- 
graphical Note. 

" Min, JCS mtg, 19 Mar 43. 



in western New Britain. These opera- 
tions were essentially the second task of 
the directive of 2 July ig42. ls With this 
statement, the Pacific Military Confer- 
ence as such came to a close, although 
the Pacific representatives remained in 
Washington a few days longer at the 
request of the Joint Chiefs of Staff. 

Preparation of the Directive of 
28 March 1943 

The Joint Chiefs of Staff, having ap- 
proved the additional Pacific reinforce- 
ments and heard the opinions of the 
Pacific delegates, immediately accepted 
the proposal that South and Southwest 
Pacific operations in 1943 be limited 
to Task Two, and turned to considera- 
tion of new orders for Halsey and Mac- 
Arthur. 19 

Neither the limitation of operations 
to Task Two nor the inclusion of Wood- 
lark and Kiriwina was an entirely new 
idea. The Joint U.S. Strategic Commit- 
tee, commissioned by the Joint Staff 
Planners to prepare a plan for the defeat 
of Japan, in February had considered 
the means for, and limiting factors af- 
fecting, the operations planned at Casa- 
blanca, and recommended that only 
Task Two be carried out in 1943. The 
committee felt that the capture of Ra- 
baul, which could not be undertaken 
until fairly late in 1943, might interfere 
with the recapture of Burma, an opera- 
tion which was considered to be on a 
priority with the advance through the 
Central Pacific and the support of 
China. The Strategic Committee had 

18 Memo by Reps of the Pac Areas, in JCS 238/2, 
2 1 Mar 43. 

19 Min, JCS mtg, 21 Mar 43. 

also recommended capture of Woodlark 
and Kiriwina. 20 

Seizure of these islands would bring 
Rabaul and the northern Solomons 
within range of fighters and medium 
bombers, and would thus compensate 
for the absence of enough heavy 
bombers. The islands, which lie outside 
the bad weather belt that frequently 
blankets the southeast tip of New 
Guinea, would also serve as staging bases 
for the rapid switching of air units 
between the South and Southwest Pa- 
cific. In December of the previous year, 
Admiral Halsey had suggested to Mac- 
Arthur the establishment of an air base 
at Woodlark or Kiriwina, and offered 
to furnish some of the necessary troops. 
This project had the approval of Mar- 
shall and King." The seizure of Wood- 
lark and Kiriw T ina was included as part 
of Plan Elkton of 12 February, but had 

-"Memo, Secretariat JUSSC for Secretariat JPS, 
13 Feb 43, sub: Opns in S and SW Pac Areas During 
1943 and Their Relation to the Concept of Mil 
Strategy for 1943 as Set Forth in the Anfa Papers, 
with Incls A and B, attached to JPS 67/2, 4 Jan 43, 
title: Proposed Dir for a Campaign Plan for the 
Defeat of Japan, ABC 381 Japan (8-27-42), Sec 1. 
This paper is also filed as JPS 67/3, 15 Feb 43, 
title: Opns in S and SW Pac in 1943. The Casa- 
blanca Conference was held at Anfa and is often 
referred to as the Anfa Conference, although its 
code name was Symbol. The Joint U.S. Strategic 
Committee was renamed the Joint War Plans Com- 
mittee in March 1943, and should not be confused 
with the Joint Strategic Survey Committee, which 
was composed of senior officers who advised on 
broad strategic matters, 

[C I N C S W P A ] , )7 Dec 4a; COMINCH to 
COMSOPAC, 18 Dec 42; Memo, King for Marshall, 
20 Feb 43, sub: Instal of Airstrips on Kiriwina 
Island or Woodlark Island; Memo, Marshall for 
King, 22 Feb 43, same sub. All in CNO File A 16-3 
(4) No. 1, Warfare Opns, SWPA, 1943, and made 
available by Lt. Grace P. Hays, USN, of the Hist 
Sec, JCS. See also Halsey and Bryan, Admiral 
Halsey's Story, p. 154. 



been omitted from the version of Elk- 
ton which Sutherland brought to Wash- 


Although the Joint Chiefs had ac- 
cepted the delegates' proposals in prin- 
ciple, they were concerned about the 
timing of operations. They brought the 
Pacific representatives and some of the 
Joint Planners into their meeting on 
Sunday morning, 2 1 March, to help set- 
tle matters. 

The Southwest Pacific delegates ar- 
gued that lack of adequate forces would 
keep the South Pacific from beginning 
operations against New Georgia and 
southern Bougainville until after the 
Southwest Pacific had seized the Huon 
Peninsula in New Guinea, an operation 
that would take place about August. 
This sequence was approximately that 
set forth in the Elkton plan. The South 
Pacific delegates, especially Harmon, felt 
that it would be better to move against 
New Georgia before the capture of the 
Huon Peninsula. A reasonable margin 
of safety would require that enough 
strength be mustered for a drive right 
through to Bougainville after Munda's 

The views of the Southwest Pacific 
delegates on New Georgia are somewhat 
curious. At an early meeting of the con- 
ference, Rear Adm. Charles M. Cooke, 
Jr., of Admiral King's staff, had asked 
Sutherland for MacArthur's opinion on 
the operation against Munda for which 
the South Pacific was then preparing. 
Sutherland replied that his chief would 
be unable to make recommendations 
until he had been "apprised" of the 
operations, the forces involved, and the 

amount of assistance he would be ex- 
pected to contribute. 22 

Admiral King was disturbed by the 
idea of postponing action in the Solo- 
mons, for the Japanese fleet was no 
longer pinned down by the Guadal- 
canal Campaign. If the Solomon opera- 
tions were to be postponed, he suggested, 
the American fleet units assigned to the 
South Pacific might be more profitably 
employed elsewhere, perhaps in the Cen- 
tral Pacific. The Joint Chiefs directed 
the Joint Planners to draft a plan, but 
did not immediately attempt to decide 
on the timing of operations. 23 In the 
message the Joint Chiefs sent to Mac- 
Arthur, Nimitz, and Halsey about the 
additional reinforcements, they stated 
that "prevailing opinion" in Washing- 
ton favored launching the invasion of 
Munda after the establishment of an air 
base at Woodlark and possibly after the 
conclusion of the planned advance in 
New Guinea. 

MacArthur replied at once to express 
his vigorous opposition to what he, Suth- 
erland, and Kenney called "divergent 
action," that is, concurrent operations 

22 Notes on Vac Conf Held in Mar 43, 3d mtg, 13 
Mar 43. It is difficult to comprehend Sutherland's 
statement as reported in the official record. 
MacArthur's message regarding Wilkinson's visit to 
Brisbane indicated that an exchange of views had 
taken place. According to the COMSOPAC War 
Diary, 4 March 1943 entry, MacArthur was in- 
formed on 4 March that South Pacific headquarters 
hoped to seize New Georgia about 10 April, Halsey 
discussed the action with Sutherland and Kenney 
at his headquarters in Noumea, New Caledonia, 
and made it clear that MacArthur would not be 
asked for any assistance except for limited air ac- 
tion against the Shordand Islands. Finally, the 28 
February Elkton, which Sutherland read to the 
conference, specifically called for an invasion of 
New Georgia and made an estimate of the forces 
that would be needed. 

2:1 Min, JCS mtg, 21 Mar 43. 


Some Pacific "Planners in Conference. From left, Capt. Cato D. Glover, Jr., 
Maj. Gen. Richard J. Marshall, Lt. Gen. George C. Kenney, Lt. Gen. Richard K. 
Sutherland, Rear Adm. Forrest P. Sherman, and Maj. Gen. Stephen J. Cham- 
berlin. Photograph taken in Brisbane, Australia, March 1944. 

against New Georgia and New Guinea 
by the South and Southwest Pacific 
Areas. Neither area, he asserted, would 
be strong enough for independent ac- 
tion. The South Pacific would need 
strong air support from its neighbor 
in the New Georgia action, and there 
simply were not enough planes. He 
therefore recommended that the New 
Georgia invasion be postponed at least 
until the seizure of the Lae-Madang 
area guaranteed control of the Vitiaz 
Strait between the Huon Peninsula and 
western New Britain, prevented the 
Japanese from moving reinforcements 
to Rabaul from the west, and enabled 
the Southwest Pacific to support and 
protect its neighbor by bombing Rabaul 

heavily. Then New Georgia could be 
taken, and the South and Southwest 
Pacific Areas, now mutually supporting, 
could begin the reduction of Rabaul. 24 
The question of timing was never 
finally determined by the Joint Chiefs. 
Speaking at their meeting on 28 March 
when the Joint Planners' draft of a new 
directive for Halsey and MacArthur was 
considered, King again emphasized the 
dangers of allowing the South Pacific to 
stand idly by while waiting for the 
northeast New Guinea coast to be 

!i Rad from MacArthur, 25 Mar 43, CM— IN 
13461; Comments by Gen Kenney on draft MS of 
this volume, attached to Ltr, Gen Kenney to Maj 
Gen Albert C. Smith, Chief of Mil Hist, 11 Nov 
53, no sub, OCMH. 



cleared. Marshall, whose talents included 
great skill at reconciling divergent 
points o£ view, offered the opinion that 
both MacArthur and Halsey would take 
every advantage to press forward when- 
ever Japanese resistance was weak. The 
Japanese would find themselves in a dif- 
ficult position. If they strengthened the 
Solomons at the expense of New Guinea 
MacArthur could move forward, and 
Halsey could take advantage of any shifts 
of troops to New Guinea. Halsey him- 
self, although willing to postpone the 
advance against New Georgia until after 
Woodlark and Kiriwina were taken, had 
stated that he would not remain idle. 
He intended to hit the Japanese with 
land-based aircraft and to be prepared 
to move into New Georgia and southern 
Bougainville if the Japanese weakened 
their defenses to such an extent that he 
could advance without precipitating a 
major engagement. King withdrew his 
objections, subject to the proviso that 
MacArthur submit detailed plans show- 
ing timing and sequence of operations 
and the composition of task forces. 25 


One final question, command, re- 
mained to be decided before the Joint 
Chiefs could issue a new directive. This 
question was settled fairly quickly. 28 The 
directive of 2 July 1942 had provided 
that Tasks Two and Three would be 
under MacArthur's direction. This prin- 
ciple continued to be accepted by the 
Joint Chiefs without serious challenge. 
Both the Army and the Navy had been 

B Min, JCS mtg, 28 Mar 43; COMSOPAC War 
Diary, 28 Mar 43 entry. 

"But see Rad, Sutherland to MacArthur, 25 Mar 
43, CM-OUT 9499. 

arguing somewhat heatedly over the 
question of a unified command for the 
entire Pacific, but the warmth of their 
debate did not seriously interfere with 
the preparation of the new directive. 
The possibility of mutual co-operation 
by Halsey and MacArthur was rejected. 
Some naval officers, including King, sug- 
gested that since Halsey would be oper- 
ating west of the line of demarcation 
(longitude 159 east), it should be moved 
westward again, but did not press the 
point. 27 It was agreed that MacArthur 
would command the operations by the 
Southwest Pacific forces, and that Hal- 
sey's operations with South Pacific forces 
in the Solomons would be under Mac- 
Arthur's "general directives." 

One particularly important aspect of 
the command question related to the 
Pacific Fleet units that would take part 
in the operations. Admiral King always 
opposed any tendency to break up the 
Pacific Fleet by permanently assigning 
its units to any particular area, for then 
the fleet would lose part of its striking 
power as well as strategic and tactical 
mobility. For these reasons King had 
previously proposed that Nimitz' au- 
thority be extended to include the waters 
of the Southwest Pacific Area, but had 
apparently never insisted on this as a 
solution. The Joint Chiefs settled the 
matter on 28 March by agreeing that all 
units of the Pacific Ocean Areas other 
than those assigned by the Joint Chiefs 
to task forces engaged in the operations 
would remain under Nimitz' general 
control. This meant that MacArthur 

" Kenney, General Kenney Reports, p. 213, indi- 
cates that discussions of this point in and out of 
the conference room were heated. The minutes do 
not yield much information about the emotions of 
the protagonists. 



would exercise strategic direction only 
over Halsey's forces that were engaged 
in the Solomons west of longitude 159° 
east, and that Halsey's other forces, as 
well as Pacific Fleet units not assigned 
by the Joint Chiefs, would remain under 
Nimitz. 28 With the question of command 
settled and the problem of timing left 
largely to the commanders' discretion, 
the Joint Chiefs on 28 March approved 
a directive providing for offensive oper- 
ations by MacArthur and Halsey in 
1943- 29 

The 28 March Directive 

Brief crisp orders were dispatched to 
Halsey, Nimitz, and MacArthur on 28 
March. The Joint Chiefs canceled their 
directive of 2 July 1942. They ordered 

28 The command question is treated in the fol- 
lowing documents: Ltr, COMINCH-CNO [King] 
to CofS USA [Marshall], 6 Jan 43, no sub, in- 
cluded in JCS 112/1, 14 Oct 42, title: Mil Sit in 
the Pac, ABC 370.26 (7-8-42), Sec 1; Memo, Gen 
Handy for Capt Connolly, Naval War Plans Div, 
29 Dec 42, no sub, OPD 384 PTO (12-29-42), Sec 
2, Case 43; Memo, Marshall for COMINCH, 8 Jan 
43, sub: Strategic Dir of Opns in the SW Pac, same 
file; Memo, COMINCH for CofS USA, 18 Feb 43, 
sub: Opns in SOPAC-SWPA, OPD Exec Off File 
No. 10, Item 67c; Memo, CofS USA for CNO, 19 
Feb 43, same sub, same file; JCS 238/3, ai Mar 43, 
title: Plan of Opns for Seizure of Solomon Islands- 
New Guinea-New Britain-New Ireland Area; JCS 
238/4, 27 Mar 43, title: Plan of Opns for Seizure 
of Solomon Islands— New Guinea-New Britain— New 
Ireland Area— Often Opns in the S and SW Pac 
During 1943; JCS 238/5/D, 28 Mar 43, title: Dir— 
Plan of Opns for Seizure of Solomon Islands-New 
Guinea— New Britain— New Ireland Area; and Min 
JCS mtg, 28 Mar 43. 

39 Min, JCS mtg, 28 Mar 43. 

MacArthur and Halsey to establish air- 
fields on Woodlark and Kiriwina, to 
seize the Lae-Salamaua-Finschhafen- 
Madang area of New Guinea and occupy 
western New Britain, and to seize and 
occupy the Solomon Islands as far as 
southern Bougainville. The operations 
were intended to inflict losses on the 
Japanese, to deny the target areas to 
the enemy, to contain Japanese forces 
in the Pacific by retaining the initiative, 
and to prepare for the ultimate seizure 
of the Bismarck Archipelago. As previ- 
ously indicated, operations would be 
conducted under MacArthur's com- 
mand. The advances in the Solomons 
were to be under the direct command 
of Halsey, who would operate under 
MacArthur's strategic direction. Except 
for those units assigned by the Joint 
Chiefs of Staff to task forces engaged in 
these campaigns, all elements of the Pa- 
cific Ocean Areas would remain under 
Nimitz. MacArthur was directed to sub- 
mit detailed plans including the com- 
position of task forces and sequence and 
timing of operations. 30 

With this directive, the Joint Chiefs 
set the program for 1943 in the South 
and Southwest Pacific. There can be no 
doubt that they were disappointed by 
their inability to approach the goals set 
so freely at Casablanca, but the 28 March 
directive possessed the virtue of being 
based on assumptions that were realistic, 
even pessimistic. The defined objectives 
were believed to be surely attainable. 

30 JCS 238/5/D, 28 Mar 43. 


Elkton III: The Plan for Cartwheel 

The Southwest Pacific Area 

Command Structure 

Most of the commands of the South- 
west and South Pacific Areas which 
would execute the Joint Chiefs' orders 
were already in existence. 1 General Mac- 
Arthur, as Allied Commander in Chief, 
had organized his General Headquar- 
ters (GHQ), Southwest Pacific Area, on 
U.S. Army lines. Directly under Suther- 
land, the Chief of Staff, 2 were the four 
standard general staff and three special 
staff sections. Each section was headed 
by an American Army officer, Officers 
from the American Navy and from the 
Australian, Netherlands, and Nether- 
lands Indies armed forces served in the 
most important staff sections, but in 
comparatively junior positions. 3 On the 

1 For details see Louis Morton's forthcoming 
volumes on strategy, command, and logistics in the 
Pacific, and Milner, Victory in Papua. For South 
Pacific organization, see below, pp. 1 67-70,! 

1 Sutherland, a lean, spare, dedicated man, and 
an exacting taskmaster, was somewhat less than 
popular with some of the officers who commanded 
forces directly under GHQ, apparently because they 
felt that he, personally, tried to take over part of 
their authority. But his worst enemies have never 
questioned his professional competence. 

3 G-g, for example, contained a substantial num- 
ber of U.S. Navy and Allied officers, and such sub- 
ordinate sections of G-2 as the Allied Intelligence 
Bureau, the Allied Translator and Interpreter Serv- 
ice, and the Allied Geographical Section had large 
numbers of Allied officers. 

surface GHQ was a U.S. Army head- 
quarters, but its responsibilities and au- 
thority were joint and Allied in nature. 
It was an operational headquarters. 

Under GHQ in Australia were three 
other tactical headquarters— Allied Land 
Forces, Allied Naval Forces, and Allied 
Air Force s, whose n ames indicate their 

(Chart 1) 

Allied Land Forces 
was commanded by an Australian, Gen- 
eral Sir Thomas Blarney, and was theo- 
retically responsible for the tactical di- 
rection of all Allied ground troops, less 
certain antiaircraft units which were 
controlled by Allied Air Forces. Under 
Allied Land Forces was the U.S. Sixth 
Army, established in the area in Febru- 
ary 1943 under command of Lt. Gen. 
Walter Krueger. Included in Sixth 
Army were Lt. Gen. Robert L. Eichel- 
berger's I Corps, the 2d Engineer Spe- 
cial Brigade, and the 503d Parachute 
Infantry Regiment. The 1st Marine Di- 
vision was under Krueger's operational 
control. 4 The First and Second Aus- 
tralian Armies, many of whose units 
were still in training, were part of Allied 
Land Forces. The main tactical head- 
quarters which operated under Blarney 

1 GHQ SWPA GO 17, 16 Feb 43, in GHQ SWPA 
G-g Jnl, 16 Feb 43. The antiaircraft units, two 
antiaircraft coast artillery brigades that were con- 
trolled by Allied Air Forces, were assigned to Sixth 

Chart 1— Organization of Forces for Cartwheel 

Gen MacArthur 

South Pacific 
Adm Halsey 

Allied Air 

Gen Kenney 

Allied Naval 
Adm Carpender 

ALAMO Force 
Gen Krueger 

U. S. Army Forces, 
Far bast 
Gen MacArthur 

New Guinea Force 
Gen Herring 

U. S. Fifth Air 

Gen Kenney 

Air Vice Marshal 




during early 1943 was New Guinea 
Force, a largely Australian headquarters 
responsible for the conduct of opera- 
tions in New Guinea. GHQ usually 
established a temporary advanced eche- 
lon at Port Moresby, New Guinea, 
shortly before the beginning of each 

Allied Naval Forces was commanded 
by Vice Adm. Arthur S. Carpender (in- 
evitably called "Chips") of the U.S. 
Navy, and included the U.S. Seventh 
Fleet (concurrently commanded by Car- 
pender) and large parts of the Australian 
and Netherlands Navies. The most im- 
portant component of Carpender's com- 
mand was the VII Amphibious Force, 
organized under Rear Adm. Daniel E. 
Barbey in early 1943. 

Lt. Gen. George C. Kenney, an Amer- 
ican airman, led the Allied Air Forces 
which consisted of the U.S. Fifth Air 
Force and the Royal Australian Air 
Force Command, Allied Air Forces, 
under Air Vice Marshal William D. 
Bostock. Kenney also commanded the 
Fifth Air Force but for tactical purposes 
it was run by the Deputy Commander, 
Brig. Gen. Ennis C. Whitehead who led 
the Advanced Echelon at Port Moresby, 
New Guinea. 5 

All national forces serving under these 
tactical headquarters were administered 
and usually supplied by their own serv- 
ice elements. U.S. Army Forces, Far 
East, commanded by MacArthur, was 
responsible for administration of the 
Sixth Army, the Fifth Air Force, and 
U.S. Army Services of Supply, South- 
west Pacific Area. This last, under Maj. 
Gen. Richard J. Marshall, had the re- 

5 Craven and Cate, The Pacific: Guadalcanal to 
Saipan, p. 99. 

sponsibility for logistical support of 
American ground forces. 6 Australian 
Line of Communications units in Allied 
Land Forces supplied the Australian 
troops. Soldiers fighting in New Guinea 
under New Guinea Force were supplied 
by a U.S.-Australian organization known 
as the Combined Operational Service 
Command which had been created dur- 
ing the Papuan campaign. 

Most echelons subordinate to GHQ 
had functioned during the Papuan cam- 
paign and by mid- 194 3 were operating 
with an efficiency born of this experi- 


The forthcoming campaigns would be 
fought in New Guinea, the Solomon 
Islands, and the Bismarck Archipelago. 
Many places in these islands bear the 
names of outstanding figures in the his- 
tory of exploration: Torres Strait, Dam- 
pier Strait, Bougainville, and D'Entre- 
casteaux Islands. Other names like New 
Britain and New Ireland are of more 
pedestrian origin, and the Bismarck and 
Solomon Seas were named during World 
War II. 7 Despite the familiarity of many 
place names, the area was one of the 
least known and least developed in all 
the world. Further, although there is 
perhaps no ideal place to fight a war, 
the New Guinea-Bismarcks-Solomons 

6 GHQ SWPA Stf Memo 3, 19 Feb 43; USAFFE 
GO i, 26 Feb 43, adv copy. Both in GHQ SWPA 
G-3 Jnl, 19 Feb 43. 

7 Nothing is named for Meneses, who first visited 
New Guinea in 1526, nor for Mendana who dis- 
covered the Solomons in 1568. But the whole group, 
the islands of Guadalcanal, San Cristobal, Santa 
Isabel, and Florida, and Point Cruz on Guadalcanal 
and Estrella Bay at Santa Isabel retain the names 
given by Mendana. 



area was o ne of the worst possible places. 
(Map 3 )\ 

All the islands have much in common, 
and much that is common is unpleas- 
ant. All have hot, wet, tropical climates. 
All are mountainous. All are heavily 
jungled. All are pest-ridden and full of 
tropical diseases, especially malaria. 
None has motor roads longer than a few- 
miles. There are almost no ports with 
piers and quays to accommodate large 

The native inhabitants are Melane- 
sians, most of them barely beyond the 
Stone Age. Cannibalism and headhunt- 
ing were suppressed only recently in 
areas where British, German, Dutch, 
and Australian governments exerted 
their authority. During World War II 
there were rumors that some of the New 
Guinea natives, freed by the Japanese 
conquests from the white man's restrain- 
ing influence, had reverted to their an- 
cient practices. 

New Guinea, the largest island in the 
area and after Greenland the largest 
island in the world, is about 1,600 stat- 
ute miles long, 500 miles from north 
to south at its widest point, and has 
an area estimated at about 312,000 
square miles. Its most distinctive geo- 
graphic feature, aside from the jungle, 
is the great cordillera that runs the 
length of the island. This cordillera 
consists of a number of parallel east- 
west mountain ranges which narrow into 
the Owen Stanley Range in the Papuan 
peninsula. The highest peaks reach over 
sixteen thousand feet into the sky. The 
mountain valleys that are cut by such 
rivers as the Sepik, Ramu, Markham, 
and Bulolo are several thousand feet 
above sea level, and the climate is pleas- 

ant and relatively healthful. There are 
no really large rivers in New Guinea, 
but the Markham, which flows into 
Huon Gulf, and the Sepik and Ramu 
are several hundred miles long. The 600- 
mile Sepik, flowing between the Victor 
Emmanuel Range and the Torricelli 
Mountains, is navigable by steam launch 
for 300 miles above its mouth. Between 
the mountains and the sea are swampy 
lowlands and vast stretches of tropical 
rain forest so thick that the sun never 
penetrates the treetops to dry the ground 
and no underbrush ever grows. 

At the outset of the war there were 
no motor roads of any significant length. 
There were short roads in and around 
the main ports and gold fields and in- 
numerable native footpaths, or "tracks." 
As both Allied and Japanese forces had 
demonstrated during the Papuan cam- 
paign, overland travel was fantastically 
difficult. The best ways to travel were 
by water and by air. However, both the 
Australians and Japanese were, in the 
first part of 1943, engaged in ambitious 
transmontane road-building projects. 

Before the war the Australians had 
exploited air travel to the utmost in 
developing the gold fields of the Bulolo 
Valley in the mountains southwest of 
Salamaua. They had avoided the dim- 
culties of overland travel by cutting air- 
strips in the flatlands of the valley, then 
flying in gold-mining machinery, build- 
ing materials, and, to add to the ameni- 
ties of life in the attractive uplands, 
even race horses. 

Across Vitiaz and Dampier Straits 
from New Guinea's Huon Peninsula 
lies Cape Gloucester, the western tip 
of New Britain, which curves northeast- 
erly to culminate in Gazelle Peninsula 

I45 c 



St Matthias I Emirau I 

I55 c 







3000 6000 9000 AND ABOVE 

50 I00MILES 

1 1 t* ■ ■■ i ■ ■ | 

— I ' 1 

50 50 100 KILOMETERS 

Feni Is 

^^Green Is 

Buka I 

Buka Passage 


L M O N 



Vella LavellifL f s~\ Kolombangara ■% 






(/^X^oodlark I 

Russell Is "s> Florida 

io° - 


e Say 


MAP 3 




Printed by Army Map Service F- Tern pie 



and Rabaul. New Ireland, long and nar- 
row, parallels the long axis of the Pa- 
puan peninsula so that it, the Admiralty 
Islands, part o£ New Guinea, and New 
Britain enclose the Bismarck Sea. New 
Britain, New Ireland, the Admiralties, 
and other islands form the Bismarck 

Southeast of Rabaul, and northeast 
of the Papuan peninsula and the Lou- 
isiade Archipelago, lie the Solomon 
Islands. This Goo-mile-long double- 
chained archipelago was found by Men- 
dana in 1568, but his calculations of 
longitude were so far wrong that two 
hundred years went by before white men 
found it again. Carteret, Bougainville, 
Surville, Shortland, and D'Entrecast- 
eaux sighted or visited the archipelago 
between 1767 and 1793, and French 
geographers eventually concluded that 
these were the islands Mendana had 

The area was divided politically. That 
part of New Guinea west of longitude 
141 east belonged to the* Netherlands. 
Papua was an Australian possession with 
the status of Territory. Northeast New 
Gui nea, the Admiralties, New Britain 
New Ireland, Bougainville, and Buka 
made up the Australian Mandated Ter- 
ritory of New Guinea; Australia took 
them from Germany in World War I 
and was awarded a League of Nations 
mandate over them.* The Solomons 
southeast of Bougainville are, politically, 
the British Solomon Islands Protector- 
ate, established by Great Britain in 1893 
to suppress blackbirding. 

8 In 1947 the Mandated Territory and Papua 
were consolidated as a United Nations trusteeship. 

A crude comparison may give a gen- 
eral idea of the size of the area. If a 
map of the New Guinea-Bismarck 
Archipelago-Solomon Islands area is 
superimposed on a map of the United 
States, with the western tip of New 
Guinea's Vogelkop Peninsula at Seattle, 
Washington, Milne Bay at southeastern 
New Guinea lies in Colorado, and the 
Solomon Islands lie in the Missouri and 
Mississippi Valleys. 


In early 1943 the key points of this 
huge area, except for Port Moresby, 
Milne Bay, Goodenough Island in the 
D'Entrecasteaux group, and the Guadal- 
canal-Russells-Florida area of the Solo- 
mons were in Japanese hands, but Al- 
lied intelligence agencies were able to 
keep a fairly close check on enemy troop, 
ship, and plane movements by means of 
radioed reports from observers operat- 
ing behind the enemy lines. These ob- 
servers were the coastwatchers, members 
of an organization, the Coastwatching 
Service, established before the war as 
part of the Directorate of Intelligence, 
Royal Australian Navy. Their territory 
originally embraced New Guinea, the 
Bismarck Archipelago, and the Solo- 
mons, but later islands of the Nether- 
lands Indies were added to the network. 
Initially the coastwatchers were all Brit- 
ish, Australian, or New Zealand civil 
servants or planters, commissioned in 
the Australian armed forces, but as the 
war progressed qualified men from the 
American forces were also assigned. The 
coastwatchers were part of the Allied 
Intelligence Bureau of the G-a Section 
of GHQ. Those in the Solomons re- 



ported their observations directly to 
South Pacific agencies. 9 

These intrepid men were greatly 
aided in their work by the devotion 
and help of the natives. The Melane- 
sians in general remained loyal to the 
Allied cause, and throughout the war 
rescued shot-down airmen and stranded 
sailors, worked as guides, bearers, and 
laborers, and a select few stayed with the 
various coastwatchers. 

As the interior of the New Guinea- 
Bismarcks-Solomons area was little 
known and practically unmapped, the 
coastwatchers proved an invaluable 
source of information on terrain. In 
addition, their hideouts served as bases 
for the patrols that thrust behind the 
Japanese lines in advance of nearly every 
Allied operation. 

The Plan of Maneuver 

On receiving the instructions from 
the Joint Chiefs of Staff, General Mac- 
Arthur and his subordinates turned to 
the job of preparing plans and issuing 
orders to carry out the directive of 28 
March. The task was not difficult. All 
that was needed was a revamping of the 
two previous Elkton plans. 

MacArthur's headquarters issued its 
plan for South and Southwest Pacific 
Area operations for 1943 on 26 April, 
and followed it the next month with 
warning orders and operations instruc- 
tions. The warning orders covered the 

"See Comdr. Eric A. Feklt, RAN, The Coast- 
watchers (Melbourne and New York: Oxford Uni- 
versity Press, 1946); MIS GHQ FEC, The Intelli- 
gence Series, IV, Operations of the Allied Intel- 
ligence Bureau, GHQ, SWTA. The Royal New 
Zealand Navy also operated a coastwatching system 
east of the Solomons. 

whole Elkton plan; the operations in- 
structions dealt specifically with the 
opening phases. The 26 April plan, 
designated Elkton III, was issued after 
a personal conference in Brisbane be- 
tween Admiral Halsey and General 

This was the first meeting of the well- 
known admiral and the even more fa- 
mous general. Halsey was deeply im- 
pressed by MacArthur; speaking of their 
wartime conferences, he wrote: 

I have seldom seen a man who makes a 
quicker, stronger, more favorable impres- 
sion. . . . On the few occasions when I 
disagreed with him, I told him so, and we 
discussed the issue until one of us changed 
his mind. My mental picture poses him 
against the background of these discussions; 
he is pacing his office, almost wearing a 
groove between his large, bare desk and 
the portrait of George Washington that 
faced it; his corncob pipe is in his hand (I 
rarely saw him smoke it) ; and he is mak- 
ing his points in a diction I have never 
heard surpassed. 10 

At this meeting, timing and co-ordi- 
nation of the advance in New Guinea 
with the invasion of New Georgia were 
discussed. 11 Halsey carried some of his 
points with MacArthur; they agreed that 
the initial invasion of New Georgia 
would take place at the same time as 
the seizure of Woodlark and Kiriwina 
instead of after the establishment of 
Southwest Pacific forces on the Huon 
Peninsula, as the Southwest Pacific lead- 
ers had been advocating. Elkton III 
specified that the New Georgia and 
Woodlark-Kiriwina operations would 

111 Halsey and Bryan, Admiral Halsey's Story, p. 

11 COMSOPAC War Diary, 25 Apr 43 entry; Hal- 
sey and Bryan, Admiral Halsey's Story, pp. 154-55. 



be simultaneous, but that major forces 
were not to be committed. 13 

The CARTWHEEL Operations 

The plan of maneuver decided on 
by MacArthur was the same as that set 
forth in previous plans— mutually sup- 
porting advances along two axes, con- 
verging finally on Rabaul. The general 
concept underlying these operations 
characterized most Allied operations in 
both South and Southwest Pacific Areas. 
Despite its stiff brand of English, the 
warning instruction that expressed this 
concept is worth noting: 

The general scheme of maneuver is to 
advance our bomber line towards Rabaul; 
first by improvement of presently occupied 
forward bases; secondly, by the occupation 
and implementation of air bases which can 
be secured without committing large forces; 
and then, by the seizure and implementa- 
tion of successive hostile airdromes. 

By destructive air attack soften up and 
gain air superiority over each attack objec- 
tive along the two axes of advance. Neu- 
tralize with appropriate aviation support- 
ing hostile air bases and destroy hostile 
naval forces and shipping within range. 
Prevent reinforcement or supply of objec- 
tives under attack. Move land forces for- 
ward, covered by air and naval forces, to 
seize and consolidate each successive objec- 
tive. Displace aviation forward onto cap- 
tured airdromes. Repeat this process to suc- 
cessive objectives, neutralizing by air ac- 
tion, or by air, land, and sea action, 
intermediate hostile installations which are 
not objectives of immediate attack. The 
entire movement will be covered by air 
attack on Japanese air and sea bases along 
the general perimeter BUKA-RABAUL- 

12 Halsey and Bryan, Admiral Halsey's Story, p. 
155: GHQ SWPA, Elkton III, Plan for the Seizure 
of the Lae-Salamaua— New Britain-Solomons Areas, 
26 Apr 43. 

KAVIENG-WEWAK, with the object of 
denying supply and reinforcement of ob- 
jectives under attack, 1 * 

The operations planned for Elkton 
III were lumped under the code name 
Cartwheel, and were arranged accord- 
ing to a complicated but flexible sched- 
ule that provided for about thirteen in- 
vasions or captures in eight months, and 
also provided for maximum mutual sup- 
port by South and Southwest Pacific 

Cartwheel would start with amphib- 
ious movements by the Southwest Pa- 
cific into Woodlark and Kiriwina. Si- 
multaneously the South Pacific, using 
"diversionary" and "aggressive" infiltra- 
tion, would move into New Georgia 
"and/or" Santa Isabel without commit- 
ting major forces to action. 14 

Woodlark and Kiriwina were not held 
by either belligerent. Kiriwina is 270 
nautical miles from Rabaul, and south- 
ern Bougainville is 300 miles away. 
Thus Allied fighters and medium bomb- 
ers would be brought within range of 
these enemy areas, and Allied control 
over the Solomon and Bismarck Seas 
would be intensified. During the seiz- 
ure of Woodlark and Kiriwina (desig- 
nated Operation I in Elkton III), heavy 
bombers would strike southern Bougain- 
ville, Buka, and Rabaul. The South 
Pacific would support the move by its 
ground operations in the Solomons (Op- 
eration A) in addition to providing stra- 
tegic naval support and pinning down 
Japanese aircraft in the Solomons. In 

"GHQ SWPA Warning Instns 2, 6 May 43, in 
GHQ SWPA G-3 Jnl, 6 May 43. 
34 GHQ SWPA, Elkton III .... 26 Apr. 43. 



accordance with Halsey's original sug- 
gestion, the South Pacific would furnish 
the occupation force and an air squad- 
ron for Woodlark. The timetable in- 
cluded in Elkton III allotted two 

months for Operations I and A. (Chart 2) 

When Operation A ended, South Pa- 
cific forces would not undertake any 
large-scale movements, but would con- 
tinue air and sea operations to support 
the Southwest Pacific. This area would 
execute Operation II, the seizure of 
Lae (I la), Salamaua and Finschhafen 
(lib), and Madang (He). Lae was to be 
seized two months after the initiation of 
the Cartwheel operations, Salamaua 
and Finschhafen six weeks after Lae, and 
Madang two weeks after Salamaua. The 
Madang operation, including the consol- 
idation phase, would probably require 
two months. During Operation II air- 
craft from both areas would keep strik- 
ing the Japanese in the Solomons, New 
Ireland, New Britain, and New Guinea. 
Airfields at Lae and in the Markham 
Valley behind Lae would support the 
advance against Madang as well as the 
South Pacific's thrust against southern 

Five and one-half months after the 
start of Cartwheel, and one month after 
the move against Lae, the South Pacific 
would complete the seizure of New Geor- 
gia, and move forward to capture the 
Japanese bases at Faisi in the Shortland 
Islands and Buin in southern Bougain- 
ville (Operation B). 15 Allied aircraft 

"This feature of Elkton III was not closely fol- 
lowed by Halsey. References to the New Georgia 
operations in Elkton III are rather vague. They 
probably were included after the Halsey-MacArthur 

from Woodlark, Kiriwina, and the Huon 
Peninsula would support and cover 
these movements. It was expected that 
Operation B would require six weeks. 

The next two sets of operations by the 
Southwest and South Pacific Areas would 
be practically concurrent. At the begin- 
ning of the seventh month, the South 
Pacific was to seize Kieta, a Japanese base 
on the east coast of Bougainville, and 
begin neutralizing the airfields on and 
near Buka (Operation C). In the middle 
of the sixth month, the Southwest Pa- 
cific would cross Vitiaz Strait to take 
Cape Gloucester and Arawe (Operation 
Ilia), then occupy Gasmata and neutral- 
ize Talasea (Operation II lb). With the 
New Guinea and New Britain bases in 
Allied hands, Wewak could be neutral- 
ized, and the operations against western 
New Britain could be supported. Finally, 
with the execution of Operations III and 
C, light bombers and fighters could eas- 
ily attack Rabaul and Kavieng, and the 
South and Southwest Pacific Areas could 
besrin to neutralize them in advance of 
an amphibious assault on Rabaul. This 
entire set of operations, it was estimated, 
would last for eight months. For plan- 
ning purposes, Elkton III assumed that 
the Cartwheel operations would begin 
about the first of June. 

The arrangements for mutual sup- 
port of the two areas during these opera- 
tions were more detailed and exact than 
those for the Guadalcanal and Papua 
Campaigns. Elkton III and subsequent 
orders, besides specifying the time and 
place of the Cartwheel operations, also 
provided for direct communication be- 
tween South and Southwest Pacific Areas. 
Starting on 15 May, daily operational 

Chart 2— Estimated Timing and Sequence of 

I— Woodlark-Kiriwina 









eiie base in 
Jarkhqm vaMey_ 



Establish additional 
. S"S «!e* n « n, s 

Ha— Lae 

lib — Salamaua- 

• ■■ Fen 

Establish air support 
for Madang-Gasmata 

I Overland approach 
I to Madang || c 


. ... dation 

Cape Gloucester 
Ilia — Arawe Development 

lllb — (josmata 







A— Infiltrate New Georgia, 
or Santa Isabel, or both 

B — New Georgia, 
Bum, Faisi 

._ .Consolidation 

C— Kieta 

Source: Adapted from a chart in GHQ SWPA's ELKTON III, 26 Apr 43, Plan For the Seizure ond 
Occupation of the Lae-Salamaua-Madang-Western New Britain-Solomons Areas. 

— ■ - — - - 



and intelligence summaries would be ex- 
changed. Instructions stressed particu- 
larly the necessity for a common radio 
frequency for fighter planes and a radio 
circuit connecting all major Allied air 
headquarters and bases. Beginning with 
Operations I and A, Southwest Pacific 
planes would conduct regular defensive 
reconnaissance over the Solomon and 
Bismarck Seas and the land areas west 
of longitude 155 east and southwest of 
the line Buka Passage-New Ireland. 
South Pacific aircraft would be respon- 
sible for defensive reconnaissance to the 
east and northeast of 155 east and the 
Buka Passage-New Ireland line, with a 
one-degree overlap granted to both areas. 
Offensive reconnaissance would be con- 
ducted without regard to any boundaries. 
MacArthur was to be notified well in ad- 
vance of any movements by air or sea, 
and all further arrangements for co-ordi- 
nation and mutual support would be 
made by him. 16 

Forces and Missions 

MacArthur mainly used the existing 
headquarters in his area, but set up one 
new task force, primarily American, di- 
rectly under GHQ. This organization, 
known at first as New Britain Force but 
from July on as Alamo Force, was com- 
manded by General Krueger, who re- 
tained his command of Sixth Army. 17 
Alamo Force headquarters was virtually 
the same as Sixth Army headquarters, 
and placing Alamo Force directly under 

16 GHQ SWPA, Elkton III ... , a6 Apr 43; 
GHQ SWPA Warning Instns 2, 6 May 43, in GHQ 
SWPA G-3 Jnl, 6 May 43. 

" Alamo was the code name in clear; the secret 
code name was Escalator Force. 

GHQ had the effect of removing most 
American troops engaged in tactical op- 
erations from General Blarney's control. 
The plans called for New Guinea Force, 
with General Blarney in command, also 
to operate directly under GHQ. Rough- 
ly speaking, New Guinea Force would 
conduct the operations in New Guinea 
while Alamo Force ran those in Wood- 
lark, Kiriwina, and New Britain. 18 All 
operations would be supported and pro- 
tected by Allied Air Forces and Allied 
Naval Forces. 

Logistical responsibilities would be di- 
vided between American and Australian 
supply services. General Marshall's U.S. 
Army Services of Supply was assigned re- 
sponsibility for the immediate movement 
of supplies for American ground forces 
by water (excluding naval movements) 
from rear bases in Australia to the inter- 
mediate bases at Port Moresby and Milne 
Bay, the advanced base at Oro Bay near 
Buna, and other bases when established. 
Marshall's command was to enlarge, 
stock, and operate ports and bases for the 
Alamo Force, and would be responsible 
for completing airdromes then under 
construction on New Guinea and Good- 
enough Island. Australian Line of Com- 
munications units were to move supplies 
from rear bases to Cape York Peninsula, 
Port Moresby, and Milne Bay. 

In amphibious assaults Allied Naval 
Forces would carry supplies forward 
from intermediate and advanced bases 
to the combat areas. When those areas 
w r ere secured the regular American and 
Australian supply agencies would take 
over. In addition, Admiral Carpender's 

18 GHQ SWPA Warning Instns a, 6 May 43, in 
GHQ SWPA G-3 Jnl, 6 May 43. 



command would assist in the movement 
of supplies forward from Australia. 

After some postponements caused by 
delays in assembling the troops for Kiri- 
wina, D Day for Woodlark, Kiriwina, 
Nassau Bay in New Guinea, and New 
Georgia in the Solomons was set for 30 

The Intelligence Estimate 

Elkton Ill's estimate of Japanese 
strength in the Bismarck Archipelago, 
Solomons, and New Guinea reflected the 
recent changes in Japanese strength and 
like the 28 February estimate was fairly 
accurate. With the Japanese in control of 
all sea and air routes leading from their 
rearward island fortresses to Rabaul, 
MacArthur and his staff clearly recog- 
nized that the enemy might quickly 
strengthen his forces. They expected that 
strong naval units from Truk, including 
6 battleships, 2 carriers, 3 auxiliary car- 
riers, 8 seaplane tenders, 15 cruisers, 
about 40 destroyers, and 27 submarines 
as well as about 50 merchant vessels of 
over 3,000 tons displacement might be 
made available at once. Within thirty 
days, about four divisions could arrive, 
as well as 277 airplanes and fleet units 
from the Netherlands Indies and the 
Philippines. By the end of six months, 
the Japanese in the Bismarck Archipel- 
ago—Solomons—New Guinea area might 
be able to muster ten to fifteen divisions 
and 755 aircraft, but not much more in 
the way of fleet strength. 

Just as on the Allied side, the crux of 
the matter would be shipping. The avail- 
ability of troopships would govern the 
size of the ground combat forces that 
could be sent to and maintained in the 

area. About 300,000 gross tons of ship- 
ping, half consisting of ships over 3,000 
tons, was immediately available, and to 
that 100,000 to 125,000 gross tons might 
be added. 

No mention was made of possible Jap- 
anese offensives against positions held by 
the Allies. Enemy capabilities were con- 
sidered to be entirely defensive. The Jap- 
anese were believed able to attempt the 
following: defense of Lae and Salamaua 
while reinforcing western New Britain 
and north New Guinea; air attacks 
against the Allied communication lines 
as well as in tactical support of ground 
defenses; naval interception of Allied 
amphibious movements; and diversions 
against northwest Australia and south- 
eastern Papua. 

Specifically, it was anticipated that the 
Japanese would attempt to hold Lae- 
Salamau while rushing about 25,000 re- 
inforcements to Madang and Finschha- 
fen by sea. Once the Allied offensives got 
under way, reasoned MacArthur's plan- 
ners, the Japanese would probably be 
unable to reinforce Lae or Salamaua. 
Enemy soldiers might be sent overland 
from Wewak through the Markham Val- 
ley to Lae, but would hardly be fit to 
fight on arrival. At the same time the 
Japanese could be expected to increase 
their garrisons at Cape Gloucester, Gas- 
mata, and Arawe in western New Britain. 

The enemy was expected to mount a 
maximum air effort in an attempt to 
defeat or delay the advancing Allies. 
Both daylight sorties and harassing night 
attacks would probably be used. If the 
Japanese could keep half their planes in 
serviceable condition, they could send 
out at least a hundred fighters and eighty- 
five bombers in the initial attacks. By 



draining the Solomons, they could attack 
on the second day with at least twenty- 
four fighters and ten bombers in addition 
to whatever aircraft were left from the 
first day, and by then more planes would 
be arriving from outside the area. The 
naval surface units at Rabaul could get 
to Lae in not more than eighteen hours, 

and strong forces could steam from Truk 
to Lae in a few days' time. 18 

19 G-a Estimate of Enemy Strength and Reinforce- 
ment Rate in the New Guinea-Bismarcks Area, 
Annex A to Elkton III. A map showing enemy dis- 
positions is appended to Annex A, and differs in 
certain minor respects from the order of battle 
data in the text of the annex. 


The Japanese 

Just as the Allies were determined to 
advance against Rabaul, the Japanese 
were determined to hold it, and, indeed, 
to continue the advance that had been 
checked at Guadalcanal and Buna. The 
importance imparted to Rabaul by its 
airfield sites and harbor, as well as by its 
strategic location, had long been recog- 
nized by the Japanese. Imperial General 
Headquarters' instructions of November 
1941 directed the capture of Rabaul at 
the earliest opportunity after the fall of 
Guam. 1 Rabaul supported the offensives 
against the Allied lines of communica- 
tion, and defensively was a bastion which 
would help defend the Caroline Islands, 
the Netherlands Indies, and the Philip- 
pines against attack from the south. It 
was one of the most important bases in 
the semicircular string of island fortresses 

1 Unless otherwise indicated all data on the 
Japanese in this chapter are derived from the fol- 
lowing monographs in the series Japanese Studies 
in World War II: 17th Army Operations, Vol. II, 
Monogr No. 40 (OCMH); 18th Army Operations, 
Vols. I, II, Monogrs No. 41, 42 (OCMH); Southeast 
Area Naval Operations, Vols. I— III, Monogrs No. 
48-50 (OCMH); History of the Army Section, Im- 
perial General Headquarters, 1941-45, Monogr No. 
72 (OCMH); Outline of Southeast Area Naval Air 
Operations, November 1942— June 1943, Pt. Ill, 
Monogr No. 107 (OCMH); and 8th Area Army 
Operations, Monogr No. 110 (OCMH). Monograph 
No. 1 10 is a greatly improved revision of History 
of the 8th Area Army, 1942-44, Monograph No. 37 

that stretched from Burma through the 
Indies and the Bismarck Archipelago to 
the Marshall Islands, thence north and 
northwest to the Kuriles. 2 

Japanese Command and Strategy 

By late 1942 Rabaul had been devel- 
oped into the major air and naval base 
in the Japanese Southeast Area, and was 
the site of the highest headquarters in 
that area. Although smaller than most 
Allied areas in the Pacific, the Southeast 
Area was huge. Its western boundary, as 
set on 2 April 1943, was longitude 140 
east. 3 The northern boundary ran from 
140 east just north of the Equator to a 
line drawn between Kapingamarangi in 
the Greenwich Islands to Nauru, thence 
southeast between the Fijis and Samoa. 
It thus embraced parts of both the South 
and Southwest Pacific Areas. 

Unlike the Allied areas, the Southeast 
Area did not possess a unified command. 
The highest Army and Navy headquar- 

1 For early Japanese planning see Morton, The 
Fall of the Philippines, pp. 51-61; Milner, Victory 
in Papua; Miller, Guadalcanal: The First Offensive, 
pp. 4-7. 

* From 7 January to 2 April 1943, the western 
boundary was the border of Dutch and Australian 
New Guinea— longitude 141 east. In addition to 
Japanese sources cited see U.S. Strategic Bombing 
Survey, The Allied Campaign Against Rabaul 
(Washington, 1946), pp. 10, 83. 

Chart 3— Organization of Japanese Forces, Southeast 

June 1943 


18th Army 
(Hq at 

Gen Adachi 

20th Division 

Gen Kitagiri 

6th Air 
(New Guine 



Adm jKoga 

8th Area 
(Hq at Rabaul) 

(Hq at Rabaul) 

Gen Imamura 


38th Division 
(New Britain) 


17th Army 
(Hq at 


65th Brigade 
(New Britain) 

Gen Hyakutake Gen Matsuda 

51st Division 

6th Division 

8th Fleet 
(Hq at 

11th Air 

(Hq at Rabaul) 

(New Georgia) 

Gen Nakano 

Gen Kanda 

Gen Sasaki 

8th Combined 

(New Georgia) 


7th Combined 

(Santa Isabel) 


Adm Ota 
Liaison and Co-operation 



ters co-operated closely with one another, 
but were responsible to different higher 
authorities. \(Chart 3) In charge of Army 

operations in eastern New Guinea, the 
Bismarck Archipelago, and the Solomons 
was General Hitoshi Imamura, com- 
manding the 8th Area Army with head- 
quarters at Rabaul. Imamura was respon- 
sible to the Army Section of Imperial 
General Headquarters. The naval com- 
mand was the Southeast Area Fleet or the 
Southeastern Fleet led by Vice Adm. 
Jinichi Kusaka. His immediate superior 
was the Commander in Chief of the 
Combined Fleet but on several occasions 
he seems to have dealt directly with 

By the time the Guadalcanal and Pa- 
puan campaigns ended, the 8th Area 
Army included two field armies and one 
air division. The ijth Army operated in 
the Solomons; the 18 th Army was to be 
responsible for the campaigns in eastern 
New Guinea. The 6th Air Division, with 
headquarters at Rabaul, generally op- 
erated in New Guinea under the tactical 
direction of the 18th Army. 

Under the Southeastern Fleet were the 
land-based nth Air Fleet, which oper- 
ated principally in the Solomons, and the 
8th Fleet with bases at Rabaul and in 
the Shortlands— Buin area. The 8th Fleet, 
whose strength and composition varied 
considerably, usually included cruisers, 
destroyers, submarines, transports, and 
naval base forces. An administrative 
rather than a battle fleet, its primary 
duties were patrol and escort. Large- 
scale combat operations were the mission 
of either the 3d or the Combined Fleet, 
both then at Truk. 4 

Both the 8th Area Army and the 
Southeastern Fleet had been set up in 
late 1942 when the Japanese, making 
their major offensive effort in the Solo- 
mons and still planning to drive the 
Americans from Guadalcanal, realized 
that they had to commit large forces to 
attain success. But Imperial General 
Headquarters then revised its strategy 
and decided to abandon Guadalcanal, 
evacuate the survivors, and withdraw to 
strong positions in front of Rabaul. 

Under the revised strategy, Imperial 
Headquarters decided to shift its empha- 
sis from the Solomons to New Guinea. 
A policy of "active defense" would be 
pursued in the Solomons in order to 
reinforce New Guinea and pursue an 
"aggressive offensive" there. 5 Lae, Sala- 
maua, Wewak, and Madang on New 
Guinea's north coast were specifically 
mentioned as bases to be held. Imamura 
therefore ordered Lt. Gen. Hatazo Ada- 
chi, commander of the 18th Army, to 
strengthen Lae, Salamaua, Wewak, and 
Madang. These points were valuable as 
harbors, airfield sites, or both. Lae and 
Salamaua were of great importance as 
their possessor could dominate Dampier 
and Vitiaz Straits and thus block any 
attempt to advance along the New 
Guinea coast to the Philippines or any 
other place in the Greater East Asia Co- 
Prosperity Sphere. The Japanese were 
determined not to yield "an operational 
route for the proclaimed enemy Philip- 
pines invasion." 6 These bases would also 
be necessary to the Japanese if they were 

*USSBS, The Allied Campaign Against Rabaul, 
pp. n, 43-44- 

"Southeast Area Naval Operations, I, Japanese 
Monogr No. 48 (OCMH), 55. 

" 18th Army Operations, I, Japanese Monogr No. 
41 (OCMH), 55-56. 



Vice Adm. Jinichi Kusaka 

to realize their hopes of capturing Port 

Thus the Japanese survivors of Buna 
were ordered to Salamaua, and on Ima- 
mura's orders Adachi directed more ele- 
ments of his army to move from Rabaul 
to the New Guinea bases. The 20th Divi- 
sion began moving to Wewak; the 41st 
sent elements to Madang, and part of the 
51st Division was sent to Lae and Sala- 
maua. The fixing of the west boundary 
of the Southeast Area on 7 January at 
the Dutch border apparently gladdened 
Adachi's heart. After being limited to 
the Buna region, "having suddenly ob- 
tained freedom of the operational area, 
it gave them [the 18th Army] bright and 
desirous hopes. ..." 7 

At the same time, detachments of Field 
Marshal Count Hisaichi Terauchi's 

' Ibid., p. 1 10. 

General Hitoshi Imamura 

Netherlands Indies-based Southern Army 
were occupying areas along New 
Guinea's north coast from the Vogelkop 
Peninsula to Hollandia. 

Imperial Headquarters' orders for the 
Solomons required the 8th Area Army, in 
co-operation with the Southeastern Fleet, 
to hold the central and northern Solo- 
mons. Army and Navy authorities at 
Rabaul disagreed over exactly where the 
forward defense lines should be located. 
The Army favored the Bougainville 
area, holding that it would be too dif- 
ficult to supply the islands farther south. 
The Navy insisted on New Georgia and 
Santa Isabel as outposts for Bougainville. 
Each service went its own way. The 
Army assumed responsibility for the de- 
fense of the northern Solomons. The 
Navy took over land defense of the cen- 
tral Solomons. Imamura gave to Lt. Gen. 



Lt. Gen. Hatazo Adachi 

Haruyoshi Hyakutake's ijth Army,, then 
consisting chiefly of the 6th Division, 
responsibility for Bougainville and ad- 
jacent islands. 8 

Having insisted on the necessity for 
holding New Georgia and Santa Isabel, 
naval authorities then complained that 
this responsibility placed an excessive 
demand on naval strength, and asked 
Imamura for some Army ground troops 
for New Georgia in addition to the few 
who were already there. The general, 
still invoking the difficulty of supply, was 
at first reluctant. In March the South- 
eastern Fleet sent the 8th Combined 
Special Naval Landing Force to New 
Georgia, and another, the yth, to Santa 

"Of the iyth Army units which served on Guadal- 
canal, the 35th Brigade {Kawaguchi Force or De- 
tachment) went to Burma; the 2d Division, to the 
Philippines; the 38th Division, to New Britain 
under direct control of the 8th Area Army. 

Isabel. 9 After a good deal of negotiation, 
and perhaps on orders from Imperial 
Headquarters, Imamura acceded to Ku- 
saka's requests and sent more Army 
troops to New Georgia under their own 
headquarters, the Southeastern Detach- 
ment, and some additional units to Santa 
Isabel. Both the 8th Combined Special 
Naval Landing Force and the Southeast- 
ern Detachment, as well as the Santa Isa- 
bel force, were under the tactical control 
of the 8th Fleet. 

Thus in early 1943 the Japanese were 
holding a network of mutually support- 
ing air and naval bases arranged in 
depth, running in two converging arcs 
through New Guinea and the Solomons 
to Rabaul. From the defensive point of 
view, these positions would serve to pro- 
tect Rabaul, the Netherlands Indies, and 
the Philippines. Offensively, these bases 
could support advances southward, and 
although the Japanese had decided on 
delaying action in the Solomons, they 
were determined to take the offensive in 
New Guinea. 

Japanese Offensives, January-June 1943 
The A ttack Against Wau 

The first offensive effort under the re- 
vised strategy was directed against Wau 
in the Bulolo Valley goldfields southeast 
of New Guinea's Huon Peninsula. Wau, 
the site of a prewar airfield, lies 145 air 
miles north by west of Port Moresby, 
and 25 air miles southwest of Salamaua. 
(Map 4)] Since May 1942 Wau had been 
held by a small body of Australians, 
known as the Kanga Force, who operated 

* Composed of the Kure 6th and the Yokosuka 
7 th Special Naval Landing Forces, the 8th Combined 
had been activated in Japan for service on Guadal- 
canal but did not get there before the evacuation. 



under control of the New Guinea Force. 
As the Bulolo Valley could be reached 
overland from other Allied bases only 
over mountainous, jungled, and swampy 
routes, the Kanga Force was supplied 
largely by air. It had been ordered to 
keep watch over Lae and Salamaua and 
to hold the Bulolo Valley as a base for 
harrying the enemy until he could be 
driven out of the area. 10 If the Japanese 
had been able to establish themselves 
at Wau, they could have reaped great 
gains. They could have staged aircraft 
from Madang and Wewak through Wau, 
thus bringing Port Moresby within ef- 
fective range of their fighters. 11 The 
18th Army entertained ambitious plans 
for capturing Wau and crossing the Owen 
Stanley Range to seize Port Moresby. It 
is not clear, however, whether Adachi 
intended to proceed from Wau over the 
rough trail that led from Wau to Bull- 
dog on the Lakekamu River, or to move 
against Port Moresby via Kokoda. Either 
route would have outflanked the Allied 
Bay positions that had been won in the 
arduous Papuan campaign. 

When 1 8th Army troops moved to 
New Guinea in early 1943, some went to 
Lae and Salamaua to strengthen naval 
forces already there. 12 The reinforced 
i02d Infantry Regiment was sent in a 
convoy from Rabaul to Lae during the 
first week in January. But the Allies, 

"> Milner, Victory in Papua, Chs. I, III; ALF, Rpt 
on New Guinea Opns: 23 Sep 42—22 Jan 44. 

" AAF Int Summary 74, 3 Feb 43, in GHQ SWPA 
G-3 Jnl, 2 Feb 43. 

"Interrogation of Lt Gen Hatazo Adachi, Lt Gen 
Rimpei Kato (former CofS, 8th Area Army), Lt Col 
Shoji Ota (former stf oft, 8th Area Army), and Capt 
Sadamu Sanagi (former Senior Stf Off, Southeastern 
Fleet), by members of the Mil Hist Sec, Australian 
Army Hq, at Rabaul, no date, OCMH. 

f. Temple 

MAP 4 

warned by the fact that the Japanese had 
given up their efforts to send troops to 
Buna, had anticipated that the Japanese 
might try to strengthen Lae and Sala- 
maua and were therefore attempting to 
isolate that area by air action. Allied 
planes found the convoy, bombed it, and 
sank two transports. About three fourths 
of the io2d went ashore at Lae, but half 
its supplies were lost. 

Once at Lae, the io2d was ordered by 
Adachi to seize Wau. This Allied enclave 
was connected to the north coast by sev- 
eral trails that could be traversed on 
foot. The Japanese commander at Lae, 
Maj. Gen. Toru Okabe, decided to begin 
his drive against Wau from Salamaua. 



Lt. Gen. Haruyoshi Hyakutake 

By 16 January he had gathered his at- 
tacking force there. 

The Allies, determined to prevent the 
Japanese from capturing Wau and 
threatening Port Moresby, had mean- 
while acted promptly. Headquarters, 
New Guinea Force, decided to reinforce 
Wau, and in mid-January advance ele- 
ments of the 17th Australian Infantry 

Brigade were flown from Milne Bay to 
Wau. 13 

After assembling at Salamaua, Okabe 
and the io2d Infantry made their way 
laboriously upward to the Bulolo Valley. 
They struck at Wau in a dusk attack on 
28 January and pushed through to the 
edge of the airfield. But there they were 
stopped. For the next three days Austral- 
ian soldiers of the 17th Brigade, plus 
ammunition, supplies, and two 25- 
pounder guns, were flown in by air. In 
three days troop carriers of the Allied 
Air Forces flew in 194 planeloads, or one 
million pounds. So critical was the situa- 
tion on the 29th that the first load of 
troops practically leaped from the planes 
firing their small arms. The Japanese 
pressed hard, but by 30 January acknowl- 
edged failure and began to withdraw. 14 

Having broken the enemy's attack, the 
Australians kept pressing him back 
toward Salamaua. In April the 3d Aus- 
tralian Division took over direction of 
operations and the Kanga Force was dis- 
solved. The Australians then halted 
short of Salamaua to wait until other 
Allied troops could be made ready for a 
large-scale attack against the entire 
Finschhafen-Lae-Salamaua complex. 15 

The Australians' gallant defense of 
Wau thus frustrated the last Japanese 
attempt to attack Port Moresby overland, 
and kept for the Allies an advantageous 

13 ALF, Rpt on New Guinea Opns: 23 Sep 42-33 
Jan 44; NGF OI 60, 13 Jan 43, in GHQ SWPA G-3 
Jnl, 14 Jan 43. 

"Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 136—37; Kenney, General Kenney Re- 
ports, pp. 186-87; ALF, Rpt on New Guinea Opns: 
a 3 Sep 43-23 Jan 44. 

"USSBS, Employment of Forces, p. 17; GHQ 
SWPA G-3 Opns Rpt 380, 21-22 Apr 43, in GHQ 
SWPA C-3 Jnl, 22 Apr 43. 



position which would help support later 
offensives against the Huon Peninsula. 

The Battle of the Bismarck Sea 

The Australian defense of Wau had a 
third consequence that was more far- 
reaching than even the most ebullient 
Bulolo Valley veteran (if anyone was 
ebullient after fighting in the mud, 
mountains, and heat) realized at the 
time. It helped lead to the destruction 
of an entire Japanese convoy and the 
subsequent weakening of Lae. 10 

Okabe's attacks against Wau had so 
depleted his meager force that the Jap- 
anese at Rabaul, who were determined to 
hold Lae and Salamaua at all costs, be- 
came worried. The 20th and 41st Divi- 
sions could not be spared from Wewak 
and Madang. Thus Imamura, Adachi, 
and the naval commanders decided to 
send the rest of the 51st Division in con- 
voy to Lae. They planned very carefully. 

They were well aware of the havoc 
that airplanes could wreak on troop 
transports. Guadalcanal had demon- 
strated that point, and if final proof was 
needed, Adachi had had it in the destruc- 
tion of part of Okabe's shipment in Jan- 
uary. Had it been possible for the con- 
voy to sail from Rabaul to Madang and 
land the troops there to march to Lae, 
the ships could have stayed out of ef- 
fective range of Allied fighters and me- 
dium bombers; heavy bombers, thus far 
relatively ineffective against ships, were 

" Unless otherwise indicated, this briel account is 
based on the Japanese monographs and on defini- 
tive accounts in Craven and Cate, The Pacific: 
Guadalcanal to Saipan, pp. 141-51, and Samuel 
Eliot Morison, History of United States Naval 
Operations in World War II, Vol. VI, Breaking the 
Bismarcks Barrier: 22 July 1942— j May 1044 (Bos- 
ton: Little, Brown and Company, 1950), Ch. V. 

not greatly feared. But there was no over- 
land or coastal route capable of getting 
large bodies of troops from Madang to 
Lae. It was therefore necessary to sail 
directly to Lae and thus come within 
range of fighters and medium bombers. 
The Japanese, employing almost two 
hundred planes based at Rabaul, Ma- 
dang. Wewak, Cape Gloucester, Gas- 
mata, and Kavieng, hoped to beat Allied 
planes down out of the air and to provide 
direct cover to the ships. 

But the Allies had deduced Japanese 
intentions. Ship movements around New 
Britain in late February, though not part 
of the effort to reinforce Lae, were noted 
by Allied reconnaissance planes. As a 
result air search was intensified and air 
striking forces were alerted. On 25 Feb- 
ruary General Kenney and his subordi- 
nates came to the conclusion that the 
Japanese would probably try to put more 
troops ashore at Lae or Madang. 17 

Not only were the Allies warned; they 
were also ready. By the end of February 
airfields in Papua, with those at Dobo- 
dura near Buna carrying the biggest load, 
based 207 bombers and 129 fighters. The 
Southwest Pacific had no aircraft carriers 
and few if any carrier-type planes that 
were specifically designed for attacks 
against ships. But Kenney and his sub- 
ordinates had redesigned the nose of the 
B-25 medium bomber and installed for- 
ward-firing .50-caliber machine guns so 
that the bomber could strafe the deck of 
a ship and thus neutralize all her exposed 
antiaircraft guns. Further, they had prac- 
ticed the skip-bombing technique that 
proved particularly effective in sinking 

" See Kenney, General Kenney Reports, p. 197. 
See also GHQ SWPA G_2 Est o£ Enemy Sit 343, 28 
Feb-i Mar 43, in GHQ SWPA G-3 Jnl, 1 Mar 43. 



Japanese Troop Transport Under Attack by B-25's, Battle of the Bismarck Sea. 

ships. 18 Once warned, the Allied airmen 
prepared detailed plans for striking the 
convoy and executed a full-scale rehears- 
al off Port Moresby. 

At Rabaul, 6,gi2 Japanese soldiers 
boarded eight ships. The ships weighed 
anchor about midnight of 28 February 
1943 and, with eight destroyers as escort, 
sailed out of Rabaul and westward 
through the Bismarck Sea at seven knots. 
At first bad weather— winds, mist, and 
rain— hid them from the air, but soon the 
weather began to break and Allied patrol 
planes sighted the convoy first on 1 

16 For details see Craven and Cate, The Pacific- 
Guadalcanal to Saipan, pp. 140—41; Kenney, Gen- 
eral Kenney Reports, pp. 21-22, 105, 144, 154—55, 
162, 164. Maj. Gen. Henry H. Arnold had observed 
the RAF practicing skip-bombing in 1941 and intro- 
duced it to the AAF. 

March and again the next morning off 
Cape Gloucester. As it was still beyond 
the reach of medium bombers, heavy 
bombers from Port Moresby attacked it 
in the Bismarck Sea. They sank one 
transport and damaged two others, a 
good score for heavy bombers. Survivors 
of the sunken ship, about 950 in num- 
ber, were picked up by two of the de- 
stroyers which made a quick run to Lae 
to land the men after dark. The destroy- 
ers returned to the convoy on the morn- 
ing of 3 March. During the night the 
convoy had sailed through Vitiaz Strait 
and into the Solomon Sea, tracked all the 
while by an Australian Catalina. 

But now the ships entered the Huon 
Gulf in clear daylight, and were within 
range of medium bombers. The Allied 
planes that had organized and rehearsed 



for the attack assembled over Cape Ward 
Hunt at 0930 and set forth for the kill. 
The Japanese had failed to destroy Al- 
lied air power in advance, and the con- 
voy's air cover was ineffective. Starting 
about 1000 and continuing until night- 
fall, American and Australian airmen in 
P-38's, P-3g's, P-40's, Beaufighters, 
A-20's, A-29's, Beaufort bombers, B-17's, 
B-24's and B-25's pounded the luckless 
Japanese from medium, low, and wave- 
top altitudes with resounding success. 
All remaining transports, along with 
four destroyers, sank on 3 and 4 March. 
After night fell motor torpedo boats 
from Buna and Tufi swept in to finish 
off crippled ships and shoot up survivors 
in the water. 

Of the 6,912 troops on board, 3,664 
were lost. Including those taken by de- 
stroyer to Lae, 3,248 were rescued by the 
Japanese. The sinking of eight transports 
and four destroyers in "the most devas- 
tating air attack on ships" since Pearl 
Harbor was a tremendous victory, and 
it was won at a cost of thirteen killed, 
twelve wounded, and four Allied planes 
shot down. 19 

The Japanese quickly changed their 
plans for future shipments. They de- 
cided to send no more convoys to Lae. 
Large slow ships would be sent only to 
Hansa Bay and Wewak; high-speed ships 

""Morison, Breaking the Bismarcks Barrier, p. 59. 
Allied casualty figures are from Kenney, General 
Kenney Reports, p. 206. Official communiques at 
the time, based on pilots' reports, claimed twenty- 
two ships and fifteen thousand men, and Kenney, 
in his hook and in his comments on the draft manu- 
script of this volume, claimed six destroyers or light 
cruisers sunk, two destroyers or light cruisers dam- 
aged, and from eleven to fourteen merchant vessels 
in the convoy sunk; he also included, in his total 
for the Bismarck Sea, two small merchant ships 
that were sunk at Lae and Wide Bay. According to 

and small craft would run to Finschhafen 
and Tuluvu on the north coast of New 
Britain. Small coastal craft would take 
men and supplies to Lae from Finsch- 
hafen and Cape Gloucester, and some 
men and supplies would be sent over- 
land from Finschhafen to Lae. In emer- 
gencies supplies that were absolutely re- 
quired at Lae would be sent in by high- 
speed ships or submarines. The main 
body of ground forces eventually intend- 
ed for Lae would be sent overland after 
completion of a road, already under con- 
struction, from Wewak through Madang 
to Lae. 

Road Building 

Construction of the road had been 
started in January. This most ambitious 
project involved building a truck high- 
way from Madang to Bogadjim, thence 
over the Finisterre Range and through 
the Ramu and Markham River Valleys 
to Lae. The 20th Division was given this 

In early February the Allies, having 
received reports from natives, were 
aware of enemy activity in the Ramu 
Valley. Allied intelligence deduced that 
the Japanese were interested in an inland 
route to Lae. Intelligence also mini- 
mized the danger of a serious threat, for 
it seemed unlikely that the road could 

Kenney a second enemy convoy joined the first, 
which explains the disparity. However, Craven and 
Cate, in The Pacific: Guadalcanal to Saipan, pp. 
147—50, and Morison, in Breaking the Bismarcks 
Barrier, after surveying all available enemy records, 
maintained that the convoy consisted of eight trans- 
ports and an equal number of destroyers, that 
there was no second convoy, and that eight trans- 
ports and four destroyers were sunk. 8th Area Army 
Operations, Japanese Monogr No. 110 (OCMH) 
supports them. 



be completed in time to be of much 
use. 20 

Allied intelligence was correct. The 
road-building projects were next to im- 
possible for the Japanese to accomplish. 
Their maps were poor. The routes they 
selected, especially the inland route for 
the Madang-Lae road, led them through 
disease-ridden jungles and swamps, over 
towering mountains, and up and across 
canyons and gorges. They never had 
enough machinery and what they had 
was ineffective. Their trucks, for exam- 
ple, were not sufficiently powerful to 
climb steep slopes. Their horses fared 
poorly on jungle grasses. Bridges kept 
washing away on the Madang-Hansa 
Bay road. Combat troops were unhappy 
as laborers. Dense forests hid the road 
builders from air observation, but in the 
open stretches of the Finisterre Range 
they were constantly subject to air attack. 
By the end of June the Madang-Lae 
road had been pushed only through the 
Finisterre Range. 

Lae therefore never did receive sub- 
stantial reinforcements or supplies, de- 
spite the Japanese determination to hold 
it and dominate Dampier and Vitiaz 

The I Operation 

While Japanese Army troops were 
busy building roads in New Guinea, 
the Japanese Navy had also taken a hand 
in an effort to beat the Allies. Galled by 
the admittedly crushing defeat in the 
Bismarck Sea, fully aware of the threat 

M See GHQ G-a Daily Summary of Enemy Int 
and G-a Est of Enemy Sit 317, 2—3 Feb 43, in GHQ 
SWPA G-3 Jnl, 3 Feb 43. See also Australian Mil 
Forces Weekly Int Review 28, 5-12 Feb 43, in GHQ 
SWPA G-3 Jnl, 5 Feb 43. 

that Allied air activity in the South Pa- 
cific presented to their shipments of 
troops and supplies to New Georgia and 
Santa Isabel, and concerned over their 
declining air strength, the Japanese de- 
cided to gather more planes, smash Al- 
lied air power, and attack Allied ship- 
ping in the Southeast Area. 

Japanese air strength was somewhat 
less than substantial at this time. In 
March 1943 there were only about three 
hundred planes— one hundred Army and 
two hundred Navy— in the Southeast 
Area. Rabaul frequently complained that 
Tokyo never sent enough replacements 
to replace losses. Toward the end of 
March General Imamura asked Imperial 
Headquarters for more. Headquarters 
did send more, but not enough to satisfy 
Imamura, and some planes that were dis- 
patched never arrived. For example, the 
68th Air Regiment navigated so badly 
while flying from Truk to Rabaul that 
many of its planes failed to find Rabaul 
and were lost at sea. 

Admiral Isoroku Yamamoto, Com- 
mander in Chief of the Combined Fleet, 
decided to take a hand in the attempt to 
beat the Allies out of the air. For this 
effort, given the code name / Operation, 
he sent the planes from the }d Fleet car- 
riers at Truk to join with nth Air Fleet 
planes at Rabaul, Kavieng, Buin, Buka, 
and Ballale. He took headquarters of 
both the Combined and 3d Fleets from 
Truk to Rabaul to direct the / Opera- 
tion, which involved more than three 
hundred aircraft. 

Japanese aircraft had concentrated 
against the Allied New Guinea bases in 
March, and the month had been a quiet 
one on Guadalcanal. But that the Jap- 
anese had renewed their interest in the 



Solomons was demonstrated to the Allies 
on 1 April when bombers and fighters 
struck at the Russells. Air combats raged 
for three hours as Allied fighters beat off 
the attackers, losing six of their number 
in the process. 21 

Six days later, 7 April, came the main 
phase of / Operation in the Solomons. 
It was a splendid opportunity for the 
Japanese, for there were many targets 
around Guadalcanal. A naval task force, 
having fueled at Tulagi, was steaming 
northwest en route to shell Vila and 
Munda that night. Including cargo ships, 
transports, and the task force, there were 
present about forty ships of corvette size 
or larger, and a larger number of smaller 
vessels. In addition much ammunition, 
fuel, and equipment were being stored 
on Guadalcanal in preparation for the 
invasion of New Georgia. 

To attack these lucrative targets, Ya- 
mamoto dispatched 117 fighters and 
71 bombers. Coastwatchers on New 
Georgia, counting more than 160 planes 
overhead, flashed warnings southward. 
Halsey canceled the scheduled bombard- 
ment; the task force rounded Florida 
and sped down Indispensable Strait. 
Other ships and craft started getting 
under way and most had reached open 
water when the Japanese arrived about 

While Allied bombers flew to the 
southeast to avoid the Japanese, all avail- 
able Allied fighters, seventy-six in num- 
ber, took the air to intercept. P-38's 
(Lightnings) flew on top, and beneath 
them, at various altitudes, were F4U's 
(Corsairs), F6F's (Hellcats), and P-39's 

11 Craven and Cate, The Pacific: Guadalcanal to 
Saipan, p. 212. 

(Airacobras). As the Japanese planes 
broke up into separate flights, a general 
melee ensued. The skies above the Rus- 
sells, Tulagi, and the waters between 
Guadalcanal and Florida saw violent 
combat. According to the Japanese, "re- 
sistance offered by the ten or so enemy 
Grummans [F6F's] and P-38's was beaten 
down and the attack on shipping was 
carried out." They reported seriously 
damaging most of the Allied ships, a 
claim that is as inaccurate as their state- 
ment that only ten Allied fighters tried 
to intercept. 22 They sank the New Zea- 
land corvette Moa, the U.S. oiler Ka- 
nawha, and the U.S. destroyer Aaron 
Ward, and damaged one other oiler. 
They apparently never sighted the task 
force. Seven Allied fighters and one pilot 
were lost, but the Japanese lost many 
more. 23 

Yamamoto, apparently satisfied with 
the performance over Guadalcanal, then 
turned against the Allies in New Guinea. 
On 11 April 22 bombers and 72 fighters 
struck at Oro Bay. They sank one mer- 
chant ship, damaged another so badly 
that it had to be beached, and hit an 
Australian minesweeper. Next day 131 
fighters and 43 bombers flew over the 
Owen Stanleys to hit Port Moresby. 
There were few Allied fighters on hand 
to oppose them. As he himself points out, 
General Kenney had expected the at- 
tack to hit Milne Bay and had sent most 

" Southeast Area Naval Operations, II, Japanese 
Monogr No. 49 (OCMH), 13. 

23 Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 212—13; Morison, Breaking the Bis- 
marchs Barrier, pp. 120-24. Allied airmen claimed 
thirty-nine Japanese planes downed in air combat, 
while surface ships claimed twenty-five. There were 
undoubtedly many duplications. The Japanese ad- 
mit losing twenty-one. 



of his fighter strength there. 24 Fortu- 
nately the damage was very light. Two 
days later the Japanese fulfilled Kenney's 
expectations by attacking Milne Bay, but 
they did little damage. One Dutch mer- 
chant ship was a total loss, and a British 
motorship and another Dutch ship were 
damaged. Yamamoto then concluded the 
/ Operation, which he regarded as highly 
successful, and returned the carrier 
planes to their parent units at Truk. The 
Japanese, apparently misled by optimis- 
tic pilots' reports, boast of destroying 1 
cruiser, 2 destroyers, 25 transports, and 
134 planes, while losing 42 planes them- 
selves. But actual Allied losses in the 
Solomons and New Guinea were 1 de- 
stroyer, 1 tanker, 1 corvette, 3 Dutch 
merchant ships, and about 25 planes. 
Ambush Over Kahili 

Yamamoto then decided to pay a 
morale-building visit to the Buin area. 
He, his chief of staff, and other officers 
left Rabaul on 18 April in two twin-en- 
gine bombers escorted by fighters. When 
the party reached a point thirty-five miles 
northwest of Kahili, the airdrome near 
Buin, they were jumped by eighteen 
P-38's from the South Pacific's Thir- 
teenth Air Force, which had been sent 
there for that very purpose. 

When Admiral Halsey returned to 
Noumea after conferring with Mac- 
Arthur in Brisbane, he learned that 
American intelligence officers had dis- 
covered the exact time on 18 April Ya- 
mamoto was due to reach the Buin area 
from Rabaul. Admiral Nimitz and his 
staff agreed that disposing of Yamamoto 
would advance the Allied cause, so the 
Commander, Aircraft, Solomons, was 

" Kenney, General Kenney Reports, pp. 225, sss>8- 

told to shoot him down. The eighteen 
P-38's, manned by picked pilots and led 
by Maj. John W. Mitchell, were sent on 
the mission. Taking off from Henderson 
Field on Guadalcanal, they flew low over 
the waves for 435 miles by a circuitous 
route to the interception point north- 
west of Kahili. Yamamoto's flight hove 
in sight just as its fighter escort was leav- 
ing. Mitchell's attack section, led by 
Capt. Thomas G. Lanphier, Jr., bored in 
and Lanphier made the kill. Yamamoto's 
plane crashed in the Bougainville jungle. 
He died. The other plane fell in the sea, 
but the chief of staff, whom it was carry- 
ing, survived. One American pilot was 
lost. This Lucifer-like descent of the ag- 
gressive, skillful Yamamoto, perhaps the 
brightest star in the Japanese military 
firmament, was a severe blow to the mo- 
rale of the Japanese armed forces. 25 

The Big Raid 

By early June, the Allies in the Solo- 
mons realized that the Japanese were 
again determined to accomplish what 
the / Operation had failed to do— cut 
the lines of communication to Guadal- 
canal by air action. Yamamoto fell from 
the skies believing that / had succeeded, 
but by June the enemy leaders at Rabaul 
knew that the Allies were freely building 
up supplies on Guadalcanal. On 7 June 
the Japanese inaugurated another se- 
ries of fighter-escorted bombing attacks 
against Guadalcanal. Planes from the 

B Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 813-14; Halsey and Bryan, Admiral 
Halsey's Story, pp. 153-57; Morison, Breaking the 
Bismarchs Barrier, pp. 128-29. There are some dif- 
ferences in these accounts, chiefly regarding Yama- 
moto's destination and time of arrival. Halsey and 
Bryan, and Craven and Cate say he was to arrive at 
Ballale at 0945. Morison says he was due at Kahili 
at 1145. 



Russells made the first interception that 
day. According to Allied accounts, the 
Japanese lost twenty-three fighters, four 
of them to P-40's of the No. 15 Royal 
New Zealand Air Force Fighter Squad- 
ron in its Solomons debut. Nine Allied 
planes were shot down but all pilots were 
recovered. In a second attack five days 
later, the Japanese are reported to have 
lost thirty-one planes, the Allies, six. 

By mid-June Allied reconnaissance 
planes were reporting 245 planes at Ra- 
baul, with the forward fields in the north- 
ern Solomons filled to capacity. What 
some Allied veterans of this period call 
"the big raid" on Guadalcanal came on 
16 June when a large force of enemy 
bombers and fighters, numbering over 
100 planes, flew down to attack Guadal- 
canal. 26 The coastwatchers again had sent 
their timely warnings, and 104 Allied 
fighters were ready. As in April, they in- 
tercepted promptly, the Japanese forma- 
tions broke up into smaller flights, and 
air combats raged. Whenever possible 
ship- and shore-based antiaircraft took 
the enemy under fire. The Japanese hit 
three Allied ships, two of which had to 
be beached, and did some damage to 
shore installations before they were 
driven off. Six Allied fighters were shot 
down. The number of enemy planes 
destroyed was large, although the exact 
total cannot be determined. The Allies 
claimed 98. One Japanese account ad- 
mits the loss of about 30 planes. 27 

20 USSBS, The Allied Campaign Against Rabaul, 
p. 46, says 150-60 planes attacked on 6 June, but 
16 June is apparently intended. 

"Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 21S— 19; U.S. Strategic Bombing Survey, 
The Thirteenth Air Force in the War Against 
Japan (Washington, 1946); Southeast Area Naval 
Operations, III, Japanese Monogr No. 50 (OCMH), 

Neither the / Operation nor "the big 
raid" achieved substantial results. The 
Japanese failed, partly because their ef- 
forts were brief and sporadic rather than 
long and sustained, and partly because 
Allied resistance had been vigorous and 
generally skillful. 

Japanese Strength and Dispositions, 
30 June ip<f3 

In June Japanese strategy was still 
substantially what it had been in Janu- 
ary. 28 Late in March Lt. Gen. Rimpei 
Kato, the 8th Area Army's chief of staff, 
and other officers had gone to Imperial 
Headquarters, apparently to explain 
things after the Bismarck Sea debacle. 
The result of the visit was an Army- 
Navy "Central Agreement" which was 
really a reaffirmation of the policies laid 
down earlier. The Japanese still planned 
to defend the Solomons while strength- 
ening the bases in New Guinea and the 
Bismarck Archipelago in preparation for 
future offensives, especially against Port 
Moresby. Ambitious plans for air su- 
premacy were prepared, including one 
for maintaining 641 planes (284 Army 
and 357 Navy) in the Southeast Area, 
but, as has been shown above, these were 
destined to fail. 

In April Imamura summoned his 
army commanders to Rabaul and gave 
them orders based on the Central Agree- 
ment. Instructions to Adachi emphasized 
holding Lae and Salamaua, building the 
Madang-Lae highway, and establishing 
coastal barge lines from western New 
Britain to Lae and Salamaua. In fulfill- 
ment of the policy of using naval air in 

M The basic research for this section was per- 
formed by Messrs. Stanley L. Falk and Burke C. 



the Solomons and Army air in New 
Guinea, the entire 6 th Air Division was 
told to move to New Guinea. 

In June Imamura issued more orders, 
which restated the importance of Lae 
and Salamaua. The 18 th Army was told 
to strengthen them as well as Wewak, 
Madang, and Finschhafen. Adachi was 
to regroup his forces at Lae and Sala- 
maua and prepare to capture the Allied 
outposts and patrol bases at Wau, Bena 
Bena, and Mount Hagen, and to infil- 
trate up the Ramu and Sepik River Val- 

In anticipation of the operations 
against Bena Bena and Hagen, Imperial 
Headquarters transferred the yth Air 
Division from the Netherlands Indies to 
the 8th Area Army about July, and 
shortly afterward placed Headquarters, 
4th Air Army under Imamura to co- 
ordinate operations of the two air divi- 

Imamura also developed an ambitious 
airfield construction program which in- 
volved building new fields or enlarging 
old ones. By June, too, all divisions of 
the 1 8th Army— the 20th, 41st, and 51st— 
were concentrated in New Guinea. The 
17th Army, still consisting chiefly of the 
6th Division, was in Bougainville. The 
Southeastern Detachment and the 8th 
Combined Special Naval Landing Force 
were dug in deeply in New Georgia, and 
the yth Combined Special Naval Land- 
ing Force and Army elements still held 
Rekata Bay at Santa Isabel. 

It is not possible, on the basis of 
existing information, to state positively 
just how many troops Imamura had un- 
der his command at this time. These 
figures, and those in the table of strength 
and dispositions, are guesses based on 

available enemy data: for the Solomons, 
25,000; for the New Guinea coast east 
of the Dutch border, 55,000; for the 
Bismarck Archipelago, perhaps 43,000 
ground troops, for a total of i23,ooo. 29 
In aircraft, the Japanese possessed a 
total of something over 500 planes in 
June, though some of them were usually 

a Strength tables and reports do not seem to 
have survived the war, and available wartime and 
postwar documents and accounts are inexact, con- 
tradictory, or both. The most detailed figures on 
Japanese Army strength in the Southeast Area for 
June 1943 are contained in Southeast Area Naval 
Operations, Vol. II, Japanese Monograph No. 49 
(OCMH), pp. aa— 23, but this account gives no hint 
o£ the source o£ the figures that are not supported 
by the few strength figures in Army records. For 
example, the naval account states that there were 
55,000 men holding Lae, Salamaua, Finschhafen, 
Madang, and Wewak— the iSth Army's area of re- 
sponsibility—in June 1943, but an 18th Army sta- 
tistical table indicates that the 18th Army then 
contained 80,000 men. The office of Official War 
History, Department of the Interior, of the Com- 
monwealth of Australia, believes that the figure of 
80,000 for the 18th Army is correct. Ltr, Mr. Gavin 
Long to Dr. Kent Roberts Greenfield (Chief His- 
torian, OCMH), 7 Dec 55, OCMH. Japanese docu- 
ments give no data on ground-troop strength in 
the Bismarck Archipelago for June 1943. The near- 
est date for which there are anything like valid 
figures is November 1943, and those figures are 
given in the U.S. Strategic Bombing Survey's The 
Allied Campaign Against Rabaul. Its text asserts 
that in November 1943 there were 97,870 (76,300 
Army and the remainder Navy) in the Bismarck 
Archipelago, but this statement does not seem to 
be solidly supported by the interrogations upon 
which the text is based. General Imamura is 
quoted on pages 10, 11, and 8a, as stating that of 
90,000 troops in the Rabaul area in November, 55,- 
000 belonged to his army. From whatever November 
figure is selected, the strength of the 17th Division, 
which arrived in New Britain in September and 
October of 1943. must be subtracted. Unfortunately, 
there are no figures on that division's strength. The 
figure 43,000, given in the text above, for the Bis- 
marcks was obtained by the arbitrary method of 
subtracting 12,000, a reasonable estimate of the 
strength of the ijth Division, from Imamura's 
November figure of 55,000. There are fewer prob- 
lems for the Solomons, which seem to have con- 
tained about 25,000 Japanese ground troops. 



Table i— Comparison of Allied Intelligence Estimates with Japanese 
Strength and Dispositions, Southeast Area, 30 June 1943 


Allied Estimates 

Bismarck Archipelago: 42,800 
(Rabaul, Talasea, Gasmata, 
Cape Gloucester, Tuluvu, 
Kavieng, Lorengau) 

8th Area Army Hq 

iSth Army Hq 

Elements }8th Div 

65th Brigade 

Naval Units 
New Guinea (east of Dutch 

border): 47,200/48,300 
Lae— Salamaua— Finschhafen: 

51st Div 



Saidor: 500 unidentified 
Madang: 12,000/13,000 

Adv Hq, iSth Army 

20th Div 
Hansa Bay: 5,000 

Wewak: 20,000 

41st Div 


Japanese Units and 
Approximate Strength 



8th Area Army Hq 
6th Air Div Hq 
38th Div 
65th Brig 
Naval Units 


51st Div 


18th Army Hq 
20th Div 

41st Div 
Naval Units 


Japanese Units and 
Approximate Strength 



Allied Estimates 

Aitape: 200 unidentified 
Vanimo: 500 Misc 
Solomons: 40,250/40,450 
Buka: 4,000 

6th Cav 



ijth Army Hq 

6th Div 

New Georgia: 10,550 

229th Inf 



6th Cav 

Santa Isabel: 2,000 

Choiseul; none listed 


zyth Army Hq 
6th Div 
13th Inf 
229th Inf 

1 Army Inf Bn 


1 light cruiser 
g/10 destroyers 
5 submarines 

461 planes 

1 cruiser 
8 destroyers 
8 submarines 


540 planes 

(ggo operational) 

Source: Allied estimates are from GHQ SWPA G-2 Int Bulletin, Summary Enemy Dispositions, 30 Jun 43, in 
GHQ SWPA G-3 Jul, 30 Jun 43 ; and CTF 31 Opn Plan A8-4J, 4 Jun 43, Int Annex, Off of Naval Reds and Library. 
Approximate strengths of Japanese units are from Japanese Monographs Nos. 37, 40-42, 48-50, 72, 110, and 107 
(OCMH). The Australian Official War historians regard the postwar estimates as too low. See Ltr, Long to Green- 
field, 7 Dec 55. They are, however, admittedly conservative, and based on the best Japanese data available in 

out o£ action. For example, of the 300 
planes assigned to the nth Air Fleet on 
30 June, only 225 were ready for combat 
operations. Of 240 belonging to the 6th 
Air Division, 50 needed minor attention 
and 25 required major repairs. 

Planes were given a high degree of 
tactical mobility by the large number 
of conveniently spaced air bases in the 
area. Kavieng had one field. Rabaul 
boasted four— Lakunai, Vunakanau, Ra- 

popo, and Keravat (which never 
amounted to much)— with one more un- 
der construction at nearby Tobera. In 
addition, the 8th Area Army was improv- 
ing fields, or building new ones, at We- 
wak, Hansa Bay, Alexishafen, Madang, 
Lae-Salamaua, Tuluvu, and Talasea. 
The same situation prevailed in the Solo- 
mons. Besides the New Georgia fields 
and the seaplane bases at Rekata Bay and 
Shortland-Faisi, there were fields at Ka- 



hili and Ballale in the Buin-Shortlands 
area, and another, Kara, soon to be built. 
There was one at Buka, with another, 
Bonis, under construction just across 
Buka Passage. On the east coast of Bou- 
gainville the Tenekau and Kieta strips 
were being built, apparently under or- 
ders of the 8th Area Army. 30 

The 8th Fleet, in June, had one 
cruiser, eight destroyers, and eight sub- 
marines. The potential of this fleet had 
been cut somewhat by Admiral Mineichi 
Koga, who had succeeded to command 
of the Combined Fleet. Because the re- 
capture of Attu in May was regarded as 
a direct threat to the Japanese homeland, 
he diverted 20 percent of the forces (ap- 
parently including aircraft) "in the 

50 In addition to the Japanese monographs cited 
above, see USSBS, The Allied Campaign Against 
Rabaul, pp. 11-12, 46; USAFISPA Photo Int Unit 
Periodic Rpt, Airdromes and Seaplane Anchorages, 
J" 1 43- 

course of being assigned or available for 
assignment" in the Southeast Area to the 
A leutians a nd to Saipan. 31 

compares Allied estimates of 

Table 1 

enemy strength, and dispositions of that 
strength, with an approximation of en- 
emy strength and dispositions based on 
available enemy records. It will be noted 
that Allied estimates for Japanese 
strength and dispositions throughout the 
Southeast Area were quite accurate. 

In June 1943 the Japanese still cher- 
ished ambitions toward future offensives. 
It is clear in retrospect that their re- 
sources made them capable of defensive 
action only. But, as at Guadalcanal and 
Buna, the Japanese were so skillful in 
defensive operations that Allied troops 
were faced with a long series of hard 

31 Southeast Area Naval Operations, II, Japanese 
Monogr No. 49 (OCMH), 18. 


Cartwheel Begins: 
The Southwest Pacific 

On 30 June 1943— D Day for Cart- 
wheel— Allied air, sea, and ground forces 
facing the Japanese from New Guinea to 
the Solomons were ready to attack. The 
Japanese were expecting the offensive 
but did not know just when or where it 
would come. And the Allies had deter- 
mined to compound their uncertainty 
by launching not one, but three inva- 
sions—in New Georgia, at Woodlark and 
Kiriwina, and at Nassau Bay in New 
Guinea in preparation for the Markham 
Valley-Lae-Salamaua operations. 

Plans and Preparations 

Planning for the seizure of Woodlark 
and Kiriwina (designated Operation 
Chronicle) had started at General Krue- 
ger's Sixth Army headquarters near Bris- 
bane in early May. General MacArthur 
had directed Allied Air and Naval Forces 
to support Alamo Force and had made 
Krueger responsible for the co-ordina- 
tion of ground, air, and naval planning. 1 
Krueger, Kenney, Carpender, Barbey, 

^HQ SWPA OI 33, 7 May 43, in GHQ SWPA 
G-3 Jnl, 8 May 43; CG Sixth Array, Hist of 
Chronicle Opn, in GHQ SWPA G-3 Jnl, 30 Aug 

and staff and liaison officers participated. 
Krueger's authority to co-ordinate plan- 
ning gave him a pre-eminent position; 
he was first among equals. 

Planning had not proceeded far be- 
fore a hitch developed. When Admiral 
Halsey suggested the seizure of Wood- 
lark and Kiriwina he offered to provide 
part of the invasion force, an offer that 
had been cheerfully accepted. Thus in 
midmonth Generals Harmon and Twi- 
ning and Vice Adm. Aubrey W. Fitch, 
who commanded all South Pacific air- 
craft, flew to Brisbane to discuss details 
of the transfer of forces to the South- 
west Pacific. On the way over from 
Noumea Harmon and Twining made 
an air reconnaissance of Woodlark, and 
on arriving at Brisbane offered their 
opinion that Woodlark would be of little 
use in providing air support for the 
South Pacific's invasion of southern Bou- 
gainville. But Kenney, Carpender, Brig. 
Gen. Stephen J. Chamberlin, G-3 of 
GHQ, and Brig. Gen. Hugh J. Casey, 
the chief engineer of GHQ, explained 
how difficult it would be for Kenney's 
aircraft to support that invasion without 
the additional airfield that Woodlark 
would provide. The South Pacific repre- 


MAP 5 

sentatives then agreed to go on with the 
operation, and the details whereby 
ground force units, a fighter squadron, 
naval construction units, and six motor 
torpedo boats would be transferred, and 
destroyer-transports (APD's) and tank 
landing ships (LST's) would be lent to 
the Southwest Pacific, were arranged. 2 
The invasion of the two islands was 
the first real amphibious movement un- 
dertaken in MacArthur's area. Planning 
was so thorough and comprehensive that 
the plans for movement of troops, sup 

2 Rad [apparently from Twining] to Comdr Ad- 
Von 5 AF, 16 May 43, in GHQ SWPA G-3 Jnl, 16 
May 43; Notes of Conf Between Reps of SOPAC 
and SWPA, Brisbane, 17 May 43, in GHQ SWPA 
G-3 Jnl, 17 May 43. 

plies, and equipment in amphibious 
shipping became standing operating pro- 
cedure for future invasions. 

Kiriwina, a narrow, north-south is- 
land twenty-five miles long, lies within 
fighter and medium bomber range of 
Rabaul, Buin in southern Bougainville, 
and Lae, and 60 miles from the nearest 
Allied base at Goodenough Island in the 
D'Entrecasteaux group. From Rabaul to 
44-mile-long Woodlark is 300 nautical 
miles, from Buin 225, from Lae 380, and 
from Goodenough 160. Neither island 
was occupied by the Japanese. (Map <;) 

MacArthur had ordered Allied Naval 
Forces to support the Alamo Force by 
carrying troops and supplies, destroying 
Japanese forces, and protecting the lines 



Brig. Gen. Nathan F. Twining, left, Lt. Gen. Millard F. Harmon, and Col. 
Glen C. Jamison examining a map of the South Pacific area. Photograph taken 
October 1942. 

of communication. To carry out these 
orders Admiral Carpender organized sev- 
eral task forces of which the most impor- 
tant were Task Forces 74 and 76. (Chart 
4) The first, commanded by Rear Adm. 
\. A. C. Crutchley, RN, and consisting 
of Australian and American cruisers and 
destroyers, was to destroy enemy ships 
in the Coral and Arafura Seas and be 
prepared to co-operate with South Pa- 
cific forces in the event of a major Jap- 
anese naval offensive. Task Force 76 was 
the Amphibious Force which had been 
organized in January 1943 under Ad- 
miral Barbey. Barbey's ships— 4 APD's, 
4 APC's, 12 LST's, 18 LCI's, and 18 

LCT's with 10 destroyers, 8 subchasers, 
4 minesweepers and 1 tug as escort- 
would transport and land the attacking 
troops. As ships at Kiriwina would be 
vulnerable to submarine attack, Barbey 
assigned 4 destroyers to cover Kiriwina 
until all defenses were in, and ordered 
PT boats to patrol at each island. 3 

Kenney's orders directed Air Vice 
Marshal Bostock's Royal Australian Air 

3 ANF Opn Plan 4-43, 19 May 43, in GHQ SWPA 
G-3 Jnl, 21 May 43; CTF 76 Opn Plan 1-43, 14 
Jun 43, in GHQ SWPA G-3 Jnl, 16 June 43; CTF 
74 Opn Order 2-43, 18 Jun 43, in GHQ SWPA G-3 
Jnl, 24 Jun 43; Ltr, CTF 76 to COMINCH, 1 Oct 
43, sub: Rpt on Opn Chronicle, in GHQ SWPA 
G-3 Jul. 5 Au g 43- 

Chart 4 — Southwest Pacific Organization for Woodlark-Kiriwina 

Gtn MacArthu 

ALAMO Force 

Allied Ai, 

Gen Kenney 

Fifth Ait 

Gen Kenn.y 

TG T0.1 

TF 71 
(Baled at 
f remantM 

Allied Naval 


TF 7! 
(Bated 01 

TF 7B 

Air Vice Manhal 

Southwell Pacific 
S*a Franlien 




TF 76 

Adm Boibey 

Woodlark TF 
Col Cunningham 
(ITSth Cavalry RCT) 

Kiriwinc, TF 
Col Herndon 
(ISSrhRCT— SdBn) 

Advance Echelon 

Gen Whitehead 

TG 76.1 
APO Fait Group 
(4 APD'i) 

TG 76.2 


TG T6.S 
Escort Group 
(6 DD .) 

TG 76.4 
LQ Taik Group 

TG 76.6 
APC Taik Group 
(4 APC.) 

TG 76.7 
Hq Group 

V Bomber Comd 
Gen Roger 

V Fishier Comd 
Gen Wuitimith 

1 it Air 
Taik Force 
Col Smith 

No. 9 Operational 
Air Commodore 



Force Command to protect the lines of 
communication along the east coast of 
Australia and to support the defense 
of forward bases, but assigned the sup- 
port of the Woodlark-Kiriwina opera- 
tion to the Fifth Air Force as a primary 
mission. The V Bomber Command, un- 
der Col. Roger M. Ramey, was to attempt 
the destruction of Japanese air power at 
Rabaul, using one heavy bomb group 
nightly from 25 through 30 June, 
weather permitting, and to attack Jap- 
anese ships, continue its reconnaissance 
missions, provide antisubmarine patrols 
during daylight within two hundred 
miles of the Allied bases in New Guinea, 
and render close support to the ground 
troops as needed. Since there were no 
Japanese on the islands support bom- 
bardment was not necessary. To Brig. 
Gen. Paul B. Wurtsmith's V Fighter 
Command went the main burden of pro- 
viding fighter escort and cover for con- 
voys and landing operations from the 
airfields at Dobodura, Port Moresby, and 
Goodenough Island. Wurtsmith was also 
directed to be prepared to station fight- 
ers on Woodlark and Kiriwina once the 
airstrips were ready. 

The 1st Air Task Force and No. g 
Operational Group of the RAAF, re- 
spectively commanded by Col. Frederic 
H. Smith and Air Commodore J. E. 
Hewitt, were ordered to destroy Jap- 
anese ships and aircraft threatening tht 
operation, and to provide antisubmarine 
escort and reconnaissance. No fighter 
umbrella was provided for the convoys, 
a lack which the naval commanders pro- 
tested vigorously but unsuccessfully. 
Fighter squadrons were maintained on 
ground alert at Dobodura, Milne Bay, 
and Goodenough Island, ready to fly if 

hostile aircraft attacked the shipping. 4 
The 112th Cavalry Regiment, Col. 
Julian W. Cunningham commanding, 
and the 158th Infantry, a separate regi- 
ment led by Col. J. Prugh Herndon, plus 
substantial supporting arms and services, 
had been allotted to the Alamo Force. 
Krueger organized the troops that had 
come from the South Pacific— the 112th 
Cavalry (a dismounted two-squadron 
unit serving as infantry), the 134th Field 
Artillery Battalion (105-mm. howitzers), 
the 12th Marine Defense Battalion, plus 
quartermaster, port, ordnance, medical, 
and engineer units, a naval base unit and 
a construction battalion— into the Wood- 
lark Task Force and ordered it to seize 
and defend Woodlark and build an air- 
field. 5 The Kiriwina Task Force, under 
Herndon's command, consisted of the 
158th Infantry {less the 2d Battalion), 

* AAF SWF A OI 36, 14 May 43, and Fifth AF OI 
3, 15 May 43. Both in GHQ SWPA G-3 Jnl, 15 
May 43. AdVon 5AF FO 83, 27 Jun 43; Ltr, CofS V 
Fighter Comd to CG Escalator, 22 Jun 43, 
sub: Protection of Shipg; Rad, CTF 76 to Comdr 
Seventh Fit, 23 Jun 43; Rad, CG AdVon 5 to CG 
Escalator, 24 Jun 43; Rad, CG Escalator to CTF 
76, 36 Jun 43. Last five in Sixth Army G— 3 Jnl and 
File No. 4, 23 Jun— 1 Jul 43. CG Sixth Army, Hist 
of Chronicle Opn, in GHQ SWPA G-3 Jnl, 30 Aug 
43; Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 164—65. 

The 1st Air Task Force consisted of a head- 
quarters based at Dobodura which had operational 
control of units temporarily assigned by General 
Whitehead. The additional headquarters was con- 
sidered necessary because the towering Owen Stan- 
leys rendered radio communication between Port 
Moresby and Dobodura somewhat temperamental. 
The Fifth Air Force thus had three headquarters 
as well as those of the bomber and fighter com- 
mands. See Craven and Cate, The Pacific: Guadal- 
canal to Saipan, pp. 164-63. 

5 A Marine Corps defense battalion consisted of 
antiaircraft batteries (go-mm., 40-mm., and 20-mm. 
antiaircraft guns, and searchlights) and coast artil- 
lery (155-mm. guns). A few defense battalions also 
included tank platoons. 



the 148th Field Artillery Battalion (105- 
mm. howitzers), plus additional 155-mm. 
gun units and engineer, ordnance, med- 
ical, antiaircraft, and quartermaster 
troops. It was to capture and hold Kiri- 
wina and construct an airdrome. The 
first echelon of the Wood! ark Force 
would be carried on 6 APD's, 6 LCI's, 
and 6 LST's, that of the Kiriwina Force 
on 2 APD's and 12 LCI's." 

Doctrine regarding unity of command 
and the passage of command from ground 
to naval officers on embarkation, and 
back to ground officers on landing, was 
not clearly set forth in the plans. For the 
relationship between naval and ground 
commanders, the principle of unity of 
command rather than co-operation seems 
to have been followed, but it would have 
been sounder to have prescribed the 
exact command relationships in the 

In contrast with the practice of the 
South Pacific Area, where naval doctrine 
prevailed, no air units were placed under 
naval or ground commanders. The ulti- 
mate authority common to air, naval, 
and ground units was GHQ itself. Air 
liaison and support parties, however, 
were set up at Alamo Force headquar- 
ters and at Dobodura. 

Krueger from the first had planned 
to establish Alamo headquarters at 
Milne Bay. When reconnaissance showed 
that development of the bay into a satis- 
factory base would constitute a sizable 
operation, he and his staff pitched in to 
do the job. 

Assembly of the invasion force was 

8 Escalator FO's 1 and 2, 2 jun 43, in GHQ 
SWPA G-3 Jnl, 12 and 10 Jun 43. Like Task Force 
76's plans these orders included so much detail as 
to constitute standing operating procedure. 

complicated by the fact that the Kiri- 
wina Force was scattered from Port 
Moresby to Australia. (The Woodlark 
Force had come virtually intact from 
the South Pacific, and was, except for 
naval and air elements, concentrated at 
Townsville). Movement schedules were 
carefully worked out, and the first ele- 
ments of the Kiriwina Force reached 
their staging area at Milne Bay in early 
June. It was soon apparent that assembly 
of the forces could not be completed 
before the third week in June. For this 
reason D Day for Chronicle, which 
would also be D Day for Nassau Bay 
and Neiv Georgia, had been set for 30 
June. 7 

On 20 June Krueger's Alamo head- 
quarters opened at Milne Bay, and Mac- 
Arthur and Barbey arrived shortly after- 
ward. Within a few days all elements of 
Herndon's Kiriwina Force reached the 
bay. Final training of this regimental 
combat team in loading and unloading 
landing craft and in beach organization 
was inhibited by the necessity for un- 
loading ships and developing the base. 
On the other hand the 112th Cavalry- 
men at Townsville were able to make 
good use of the opportunity to train 
uninterruptedly. Barbey's amphibious 
force, Task Force 76, was also able to 
train effectively, an activity that had 
begun in early May. 8 

At Townsville and Milne Bay, soldiers 
and sailors marked "loading slots" or 

1 Rpt o£ Com Appointed by Gen Krueger, 25 May 
43, in GHQ SWPA G-3 Jnl, 27 May 43; GHQ SWPA 
OI 33/10, 17 Jun 43, in GHQ SWPA G-3 Jnl, 8 
May 43. 

8 CG Sixth Army, Hist of Chronicle Opn, in GHQ 
SWPA G-3 Jnl, 30 Aug 43; Byproduct [Kiriwina] 
TF Jnl and Hist of Kiriwina TF; CTF 76 Rpt on 



deck-plan layouts of LST's and LCT's 
on the beaches with tape, then assembled 
loads in the slots to test the cargo space 
allotted against the cargo assigned. All 
units agreed the technique worked very 

During the last days of June bad 
weather prevented the planned air at- 
tacks against Rabaul, but B-25's and A- 
20's made about seventy sorties against 
Lae and Salamaua. On 30 June the 
weather cleared and eight B-17's and 
three B-24's attacked Vunakanau air- 
strip at Rabaul. Bombing on this small 
scale, which was all the resources in the 
area would permit, continued for the 
next few days while the ground troops 
consolidated themselves at Woodlark 
and Kiriwina. 9 

The Advance Parties 

In early May two small engineer recon- 
naissance parties headed by the Sixth 
Army's deputy engineer had slipped 
ashore on Woodlark and Kiriwina to 
gather data on airfield sites, beach con- 
ditions, and defense positions. 10 Their 
reports, coupled with the fact that there 

9 Craven and Cate, The Pacific: Guadalcanal to 
Saipan, p. 166. 

10 This and the next two subsections are based on 
Morison, Breaking the Bismarcks Barrier, Ch. IX; 
Office of the Chief Engineer, General Headquarters 
Army Forces Pacific [GHQ AFPAC] Engineers of 
the Southwest Pacific: 1941-1945, I, Engineers in 
Theater Operations (Washington, 1947), 100—102; 
CG Sixth Army, Hist of Chronicle Opn, in GHQ 
SWPA G-3 Jnt, 30 Aug 43; Ltr, CTG 76.1 to CTF 
76, 24 Jun 43, sub: Adv Landing Leathf.rback 
[Woodlark], in Sixth Army G-3 Jnl and File No. 
4, 23 Jun-i Jul 43; Ltr, Col Cunningham to CG 
Escalator, 3 Jul 42, sub: Current Opns Leather- 
back TF, in Sixth Army G— 3 Jnl and File No. 4, 2 
Jul-10 Jul 43; CTF 76 Rpt on Chronicle; Sixth 
Army G-3 Jnl and File for the period covered; 
Woodlark TF [112th Cav RCT] Opns Diary; By- 
product TF Jnl and Hist of Kiriwina TF. 

were no Japanese troops present, indi- 
cated that it would be advisable and pos- 
sible to send in parties to prepare beaches 
and roads in advance of the main land- 
ings. Thus Chronicle was unusual 
among amphibious operations, for the 
shore party landed ahead of the assault 

At 0400, 21 June, the APD's Brooks 
and Humphreys left Townsville carry- 
ing almost two hundred men of the 1 12th 
Cavalry. They stopped at Milne Bay to 
pick up more men the next day, and at 
1600 left Milne Bay at high speed to 
make the night run to Woodlark. The 
trip was timed to keep the ships within 
range of fighter cover until dusk on the 
outgoing trip, and after dawn on the 
return voyage. The APD's reached 
Woodlark without incident, and at 0032 
of 23 June the advance party, under 
Maj, D. M. McMains, started landing at 
Guasopa Harbor in six LCP(R)'s. Rough 
seas and high winds slowed the landings, 
which were not completed until 0400, 
when the APD's shoved off for Milne 

The Australian coastwatcher had not 
been informed before the landing. When 
told that troops were coming ashore 
he formed his native guerrillas in skir- 
mish line and got ready to fight. Fortu- 
nately before anything tragic happened 
he heard the invaders speaking the Amer- 
ican variety of English and joined them. 

The Brooks and Humphreys reached 
Milne Bay during daylight of 23 June 
and took aboard the 158th Regimental 
Combat Team's shore party, a part of 
the 59th Combat Engineer Company and 
the 158th Infantry's communication pla- 
toon, under command of Lt. Col. Floyd 
G. Powell. Departing Milne Bay at 1810, 



four hours behind schedule, they reached 
Kiriwina at midnight. 11 The island is 
almost entirely surrounded by a coral 
reef, with a five-mile-long channel wind- 
ing through the reef to a 200-yard-wide 
beach at Losuia on the south coast of 
the main part of the island. Unloading 
of the APD's went very slowly as the LCP 
(R)'s threaded their way through the 
channel. The tide was low, and the land- 
ing craft ran aground several times in 
the darkness. Admiral Barbey also 
blamed the i58th's inadequate training 
for part of the delay. Daylight came be- 
fore the ships were emptied; they de- 
parted with part of their loads still on 
board. Three nights later they returned 
to unload heavy communication and en- 
gineer equipment that had been left in 
their holds. This led Barbey to recom- 
mend that APD's carry no item of equip- 
ment that could not readily be carried by 
one man. 

At Woodlark the advance party recon- 
noitered, established outposts and beach 
defenses, dug wells, blasted coral ob- 
structions out of the channels, cleared 
trails and dispersal and bivouac areas, 
prepared six beaching points for LST's, 
and installed signs, markers, and lights 
to mark channels and beaches for the 
main body, which would be landing in 
darkness to avoid Japanese air attacks. 
Similar efforts by the Kiriwina party 
were not as successful, partly because of 
the delay in landing engineer equip- 
ment. A good deal of effort was expended 
in building a coral causeway, 7 feet high 
and 300 yards long, across the reef on 

11 Col Herndon's comments on draft MS of this 
vokime, attached to his 1st Ind, )6 >.'ov 53, to Ltr, 
Gen Smith, Chief of Mil Hist, to Col Herndon, 6 
Oct 53, no sub, OCMH. 

the north coast to permit a landing there. 
Natives aided in this work by lugging 
basketloads of coral. 

The Japanese were unaware of, or 
indifferent to, the advance parties; they 
launched neither surface nor air attacks 
against them. 

The Landings 

About half the Woodlark Force— units 
of the 112th Cavalry, the 134th Field 
Artillery Battalion, and the 12th Marine 
Defense Battalion— left Townsville on 
25 June aboard six LST's, with one sub- 
chaser and two destroyers as escort. The 
voyage to the target was uneventful. 
Landing of the 2,600 troops began at 
2100 of 30 June. Unloading of the LST's 
at their beaching points w r as rapid. Cun- 
ningham's force had borrowed extra 
trucks at Townsville to permit every 
item of equipment to be put aboard a 
truck which was driven aboard an LST 
at Townsville, then driven off at Wood- 
lark. Emptied of their loads, the slow- 
moving LST's cleared Woodlark before 

Two APD's, carrying part of the 
Woodlark Force from Milne Bay, ar- 
rived shortly before 0100, 1 July, but 
encountered trouble in navigating the 
channel with the result that landing 
craft were not put into the water until 
0230. The landing craft coxswains had 
trouble finding the right beach, but by 
0600 the APD's were emptied and ready 
to leave. Some confusion had existed on 
the beach, but not enough to prevent its 
being cleared by the same time. 

Additional echelons arrived in LCI's 
and LST's on 1 July, and all these were 
unloaded quickly and easily. The LST's 
took 310 instead of the 317 trucks, Cun- 



ningham explained, because one LST 
raised its bow ramp and closed its doors 
before all its trucks could be driven 

On shore, defense positions were set 
up. Antiaircraft and coast artillery pieces 
of the 12th Defense Battalion were in- 
stalled, and machine gun and 37-mm. 
beach positions were established. Cargo 
was moved inland, and work on the air- 
field began on 2 July. 

Meanwhile Colonel Herndon's Kiri- 
wina Force had been landing, but with- 
out the smoothness that characterized 
operations at Woodlark. Shortly after 
dawn on 30 June, twelve LCI's, which 
with six escorting destroyers had sailed 
from Milne Bay the previous noon, be- 
gan landing their 2,250 troops. Trouble 
accompanied the landing from the start. 
The LCI's had great difficulty getting 
through the narrow, reef-filled channel 
to Red Beach near Losuia. And the 
water shallowed near shore so much that 
they grounded 200-300 yards from the 
shoreline. The landing went slowly. 12 

Sunset of 30 June saw the arrival of 
twelve LCT's and seven LCM's which 
had left Milne Bay on 29 June and 
stopped overnight at Goodenough Is- 
land. Again there were problems. Heavy 
rains were falling. The tide was out. 
Only one LCT was able to cross a sand- 
bar which blocked the approach to the 
jetty at Losuia. Other LCT's hung up 
on the bar and were forced to wait for 
the tide to float them off. The remainder 

11 Colonel Herndon stated on 16 November 1953 
that part of the trouble arose from a navigational 
error that caused the ships to sail past Kiriwina and 
made them late. He also stated that, originally, the 
main body was to land on the north coast, but that 
for some reason the plan was changed and Red 
Beach and Losuia Jetty were used. 

Troops Disembarking From LCI at 
Kiriwina Island wade ashore, 30 June 

made for Red Beach but grounded 
offshore with the result that much of the 
gear on board had to be hand-carried 
ashore. Some of the vehicles were driven 
ashore, but several drowned out in the 
salt water. 

LCT's in subsequent echelons avoided 
some of the difficulties by landing on the 
north shore of Kiriwina where the coral 
causeway had been built. Here trucks 
could back right up onto the bow ramps 
of the LCT's, but several were damaged 
by sliding off the causeway. 

In the absence of enemy interference 
Admiral Barbey approved a change in 
the original plan to move part of the 
supplies to Goodenough aboard LST's, 
then transship them to LCT's for the 
trip to Kiriwina. After 12 July LST's 



Natives Carrying Luggage which had been deposited on the coral causeway, 
north shore of Kiriwina Island, i July 194}. 

sailed directly from Milne Bay to the 
north shore of Kiriwina. 

Unloading on the north shore, while 
easier than at Losuia, complicated mat- 
ters further for the troops ashore. Heavy 
equipment was landed some distance 
from the proposed airfield near Losuia. 
Building the necessary roads was slowed 
by heavy rains and lack of enough heavy 
engineer equipment. 

Base Development 

Meanwhile the construction program 
at Woodlark went forward. By 14 July 
the airfield was near enough completion 
to accommodate C-47's. One week later 
5,200 feet of runway were surfaced with 
coral, and on 23 July the air garrison— 
the 67th Fighter Squadron which had 

served on Guadalcanal in the grim days 
of 1942— arrived for duty. 

On Kiriwina heavy rains continued 
and added to the engineers' troubles in 
building and maintaining roads. All con- 
struction equipment was used on the 
roads until about 10 July; during that 
time the airfield site was partly cleared 
with hand tools. General Krueger visited 
the island on 1 1 July and expressed his 
dissatisfaction with the progress of road 
and airfield construction. Three days 
later he placed Col. John T. Murray, 
formerly of the 41st Division, in com- 
mand of the Kiriwina Task Force and 
returned Colonel Herndon to command 
of the 158th Infantry. Herndon had 
asked for more engineers and machinery. 
These arrived after Murray took com- 



mand and thereafter the work went 
faster. By D plus 20 the first airstrip, 
1,500 feet long, was cleared, roughly 
graded, and ready for surfacing. By the 
month's end the strip was 5,000 feet long 
and ready for coral. No. 79 Squadron of 
the RAAF flew in and began operations 
on 18 August. 

Except for reconnaissance and two 
small bombing attacks against Wood- 
lark, the enemy did not react to the in- 
vasions, so that Barbey was able to trans- 
port twenty echelons to Kiriwina and 
seven to Woodlark without losing a ship 
or a man. By mid-August transport of 
supplies and men to the two islands was 
no longer a tactical mission. U.S. Army 
Services of Supply was ready to relieve 
Barbey of logistical responsibility. 

Thus the Southwest Pacific Area, using 
small forces, was able to secure two more 
airfields to further the Allies' control 
over the Solomon Sea. 

Nassau Bay 
Plans and Preparations 

The invasion of Nassau Bay was de- 
signed to ease the problem of supplying 
the troops that were to attack Salamaua 
and Lae. They could not be wholly sup- 
plied by ship, by landing craft, by air- 
plane, or by land. The threat of Japanese 
air attacks in the restricted waters of 
Huon Gulf and Vitiaz Strait, coupled 
with the prevailing shortage of troop 
and cargo ships, rendered the use of large 
ships impractical if not impossible. The 



Clearing Airfield Site With Hand Tools, Kiriwina Island, July 1943. 

shortage of landing craft and the dis- 
tance limited the extent of any shore-to- 
shore operations. The Australian troops 
operating out of Wau against Salamaua 
were still being supplied by air, and this 
placed a heavy burden on Southwest 
Pacific air transport and limited the num- 
ber of ground troops that could be em- 
ployed. In order to supplement air trans- 
port the Australians had begun their 
road from Edie Creek at the south end 
of the Bulolo Valley to the headwaters 
of the Lakekamu River on the south- 
west coast of the Papuan peninsula, but 
the tremendous difficulties inherent in 
pushing roads through New Guinea 
mountains slowed the Australians as 
they had the Japanese. It was clear that 
the opening of the Markham Valley- 
Huon Peninsula campaign would be de- 
layed beyond August if it had to await 

completion of the mountain highway. 13 
The seizure of Nassau Bay offered a 
possibility of at least partially solving 
these problems, a possibility which fitted 
neatly into the pattern of plans already 
being prepared. Nassau Bay lies less than 
sixty miles from Lae, or within range 
of the landing craft of the 2d Engineer 
Special Brigade which GHQ expected 
to employ, and it is just a short distance 
down the Papuan coast from Salamaua. 
Troops of the 3d Australian Division 
were operating inland from Nassau Bay 
at this time. Seizure of the bay by a 
shore-to-shore movement from Morobe, 
then held by the U.S. i62d Infantry 

"USSBS, Employment of Forces, pp. 21—22; ALF, 
Rpt on New Guinea Opns: Wau-Salamaua, 22 Jan— 
13 Sep 43; Memo, Comdr ALF for GHQ SWPA, 5 
May 43, sub: Warning Instns, in GHQ SWPA G-5 
Jnl, 6 May 43. 



of the 41st Division, would provide a 
means by which the Australians getting 
ready to attack Salamaua could be sup- 
plied by water to supplement the air 
drops, and would also provide a staging 
point for the shore-to-shore movement 
of an entire Australian division to a 
point east of Lae. Therefore GHQ and 
New Guinea Force headquarters decided 
to seize Nassau Bay on the same day that 
Woodlark, Kiriwina, and New Georgia 
were invaded. The troops seizing Nassau 
Bay would then join forces with 3d 
Australian Division and press against 
Salamaua in order to keep the Japanese 
from deducing that the Allies were plan- 
ning a major assault against Lae." 

General Blarney was supposed to as- 
sume personal command of the New 
Guinea Force for the Markham Valley- 
Huon Peninsula operations but the pres- 
sure of his duties kept him in Australia 
until August. Pending his arrival in New 
Guinea Lt. Gen. E. F. Herring of the 
Australian Army retained command of 
the New Guinea Force and operated 
under Blarney's headquarters instead of 
GHQ as originally planned. Maj. Gen. 
Stanley G. Savige, General Officer Com- 
manding the 3d Australian Division, had 
tactical command of the operations 
against Salamaua. Troops of the U.S. 
i62d Regimental Combat Team, which 
was assigned to Nassau Bay and sub- 
sequent operations against Salamaua, 
would come under General Savige's con- 
trol once they were ashore. 

14 GHQ SWPA OI 33, 7 May 43, in GHQ SWPA 
G-3 Jnl, 8 May 43; Ltr, Land Hq [ALF] to Gen 
Off Commanding NGF, 17 May 43, sub: Postern— 
Seizure Lae— Salamaua-Finschhafen-Madang Area, 
in GHQ SWPA G-3 Jnl, 7 Jun 43; GHQ SWPA OI 
34, 13 Jun 43, in GHQ SWPA G-3 Jnl, 14 Jun 43. 

When the Australians had defeated 
the Japanese attempt to capture Wau, 
they pursued the retreating enemy out 
of the Bulolo Valley and down through 
the mountains to a point inland from 
Nassau Bay. In preparation for Nassau 
Bay and the attack on Salamaua, Savige 
ordered his division to push against Sala- 
maua from the west and south. He di- 
rected the MacKechnie Force, essentially 
a battalion combat team of the 16 2d In- 
fantry, to make the initial landing at 
Nassau Bay and operate on the right 
(east) flank of his 17th Brigade. At the 
same time the 24th Australian Infantry 
Battalion would create a diversion by 
operating against the Japanese detach- 
ments in the Markham Valley and estab- 
lishing an ambush on the Huon Gulf at 
the mouth of the Buang Ri ver, halfw ay 

(Map 6) 

between Lae and Salamaua. 

From 20 through 23 June the Japanese 
counterattacked the 17th Brigade's posi- 
tions in the vicinity of Mubo and Laba- 
bia Ridge, a 3,000-foot eminence that is 
surrounded by the Bitoi and Buyawim 
Rivers and has a commanding view of 
Nassau Bay to the southeast, Bitoi Ridge 
to the north, and the Komiatum Track 
which served as the line of communica- 
tions from Salamaua to the Japanese 
facing the Australians. The Japanese 
fought hard but failed to budge the 17th 
Brigade. Starting on 23 June they retired 
a short distance to the north. On 30 June 
Savige's 15th Brigade was attacking Bob- 
dubi and the 17th Brigade, facing north, 
was holding Mubo and Lababia Ridge. 1 * 

"ALF, Rpt on New Guinea Opns: Wau-Sala- 
maua, 22 Jan— 13 Sep 43; Incl 1, Tactical Sit to 1630, 
30 Jun 43, to GHQ SWPA G— 2 Daily Summary of. 
Enemy Int and G— 2 Est of Enemy Sit 465, 30 Jun— 
1 Jul 43, in GHQ SWPA G-3 Jnl, 1 Jul 43. 



MAP 6 

The MacKechnie Force, designated to 
land at Nassau Bay on 30 June, consisted 
of the reinforced 1st Battalion, i62d 
Infantry. In command was Col, Archi- 
bald R. MacKechnie, commander of the 
i62d. This regiment had arrived in New 
Guinea from Australia in February 1943. 
Organized in March, the MacKechnie 
Force moved by land marches and sea- 
borne movements in landing craft and 
trawlers from the Buna-Sanananda area 
to Morobe, where it set up defensive 
positions to protect an advanced PT boat 
base. For Nassau Bay the force was aug- 

mented by American and Australian 
units. 10 

By late June the 3d Battalion, i62d, 
had relieved the MacKechnie Force of 
the mission of defending Morobe. Thirty 
days' supply and ten units of fire had 
been assembled. The troops trained for 
the landing by boarding PT boats, then 
transferring at sea to LCVP's, and de- 
barking on beaches from the landing 
craft. On the night of 28 June the Intel- 
ligence and Reconnaissance Platoon, 
16 2d, outposted the islands lying off- 
shore between Nassau Bay and Mageri 
Point about ten miles north-northwest 
of Morobe, where the invasion was to 
be mounted, in order to install lights 
to guide the invasion flotilla. Colonel 
MacKechnie flew to the Bulolo Valley 
for a conference with General Savige, 
and at his request Savige dispatched one 
of his companies from Lababia Ridge 
to the mouth of the Bitoi River to divert 
Japanese attention from Nassau Bay. As 

18 i62d In£ Rpt of Opns, 29 Jun-ia Sep 43, in 
Morobe-Nassau— Bitoi Ridge— Mt. Tambu-Tambu 
Bay-Salamaua Area of New Guinea; William F. Mc- 
Cartney, The Jungleers: A History of the 41st In- 
fantry Division (Washington: Infantry Journal 
Press, 1948), p. 51; Ltr, Col MacKechnie to Gen 
Smith, Chief of Mil Hist, 20 Oct 53, no sub, OCMH. 

The augmented MacKechnie Force consisted of 
Lt. Col. Harold R. Taylor's 1st Battalion, i62d; 
one platoon of the regimental Antitank Company; 
part of the regimental Service Company; one com- 
pany of the 116th Engineer Battalion; elements of 
the 116th Medical Battalion and a portable surgical 
hospital; the 218th Field Artillery Battalion (75- 
mm. pack howitzers), less A Battery; detachments 
from the 41st Division signal, quartermaster, and 
ordnance companies; detachments of the Combined 
Operational Service Command and the Australian 
New Guinea Administrative Unit, a military or- 
ganization in charge of native affairs; a detachment 
of C Battery, 209th Coast Artillery Battalion (Anti- 
aircraft); A Company, Papuan Infantry Battalion 
(native enlisted men and Australian officers); and 
A and D Companies of the 532d Engineer Boat and 
Shore Regiment, 2d Engineer Special Brigade. 



the landing was to be made in darkness, 
one platoon of this company was sent to 
the landing beach to set up lights to 
guide the landing craft. Company A, 
Papuan Infantry Battalion, of the Mac- 
Kechnie Force, reconnoitered to Cape 
Dinga just south of Nassau Bay, and one 
of its scouts even sneaked into the enemy 
camp at Cape Dinga and spent the night 
with the Japanese. On the basis of the 
Papuan Infantry Battalion's reports it 
was estimated 300-400 Japanese were 
in the vicinity of Nassau Bay, and about 
75 more near the south arm of the Bitoi 
River. 17 

The Enemy 

This estimate was somewhat exagger- 
ated. Present at Cape Dinga were about 
a hundred men of the iozd Infantry, 51st 
Division, and about fifty sailors of a naval 
guard unit. 18 The Japanese were expect- 
ing an Allied landing to come in Huon 
Gulf rather than at Nassau Bay, and 
had made their dispositions accordingly. 

General Adachi, commanding the 18th 
Army from his headquarters at Madang, 
had been carrying out the 8th Area Army 
commander's orders to strengthen We- 
wak, Madang, Finschhafen, and espe- 
cially Lae and Salamaua to protect Vitiaz 
Strait while preparing to attack Wau, 
Bena Bena, and Mount Hagen and infil- 

trate the Ramu and Se pik River Valleys 

" 1G21I Inf Rpt of Opns; McCartney, The Jun- 
gleers, p. 52. 

18 This subsection is based on 8th Area Army 
Operations, Japanese Monogr No. 110 (OCMH), 
pp. 43—45; 18th Army Operations, II, Japanese 
Monogr No. 42 (OCMH), 1—22; 18th Army Opera- 
tions, Annex B (Maps), Japanese Monogr No. 47 
(OCMH); Hist Div MIS GHQ FEC, Statements of 
Japanese Officials on World War II (English Trans- 
lations), IV, 119-20, OCMH; Interrogation of 
Adachi et al., by Mil Hist Sec, Australian Araiy 
Hq, OCMH. 

(See below, Map 12.) The Madang-Lae 
Highway was still under construction 
but had been pushed only to the Fin- 
isterre Range which parallels the north 
coast of the Huon Peninsula. The Jap- 
anese correctly estimated that the Allies 
planned to use the air base sites in the 
mountain valleys to support their ad- 
vances along the coast. Therefore they 
planned the moves against Wau and 
against Bena Bena and Mount Hagen, 
two outposts that had been used since 
1942. The 6th Air Division, based in the 
Wewak area, was ordered to attack these 
points daily. 

In command at Lae was Maj. Gen. 
Ryoichi Shoge, infantry group com- 
mander of the 41st Division. His com- 
mand at this time was largely transient, 
as the 1 8th Army was sending troops 
through Lae to strengthen Salamaua. 
Since the March disaster in the Bismarck 
Sea, some troops had been landed at Lae 
from submarines, forty men per boat; 
others came in barges and destroyers to 
Cape Gloucester from Rabaul, thence to 
Finschhafen by barge and overland or 
by barge to Lae. In April and May the 
66th Infantry (less the 3d Battalion), 
•yisl Division, had been transferred to 
Salamaua from Lae, and elements of the 
115th Infantry, the 14th Artillery Regi- 
ment, and the 51st Engineer Regiment, 
all of the 51st Division, staged through 
Lae for Salamaua. At Salamaua Lt. Gen. 
Hidemitsu Nakano, commander of the 
51st Division, was directing operations. 

The third infantry regiment of Na- 
kano's division, the io2d, had made the 
January attack against Wau and had 
been almost continuously in action since 
that time. 



By the end of June Nakano had six 
thousand men under his command. The 
Japanese defensive positions included the 
high ground inland from the shore- 
Mount Tambu, Komiatum, and Bob- 

Landing of the MacKechnie Force 

As dusk fell at Morobe on 29 June 
three PT boats of the Seventh Fleet took 
aboard 210 men of the MacKechnie 
Force. A fourth PT, without passengers, 
escorted. 19 At the same time twenty-nine 
LCVP's, two Japanese barges, and one 
LCM of the 532d Engineer Boat and 
Shore Regiment took the other 770 men 
of the MacKechnie Force on board at 
Mageri Point. The landing craft were 
organized in three waves which departed 
Mageri at twenty-minute intervals. The 
night was dark, the sea heavy; rain was 

The first two waves rendezvoused with 
the two PT boats from Morobe which 
were to guide them to the target but the 
third missed and proceeded on the forty- 
mile run to Nassau Bay without a guide. 

Thus far things had gone fairly well 
but the remainder of the night was full 

"This subsection is based on McCartney, The 
Jungleers, pp. 52—55; Morison, Breaking the Bis- 
marcks Barrier, pp. 136-37; Office of the Chief En- 
gineer, GHQ AFPAC, Engineers of the Southwest 
Pacific: 1941-1945, VIII, Critique (Washington, 
1951), 84-85; ALF, Rpt on New Guinea Opns: 
Wau— Salamaua, 22 Jan- 13 Sep 43; Ltr, Brig Gen 
William F. Heavey, CO 2d ESB, to Chief Engr 
SWPA, 13 Jul 43, sub: Rpt on Nassau Bay Opns, 
in GHQ SWPA G-3 Jnl, 19 Jul 43; GHQ SWPA 
G-g Jnl for period covered; 41st Div G-3 Jnl and 
File for period covered; i6ad Inf Rpt of Opns, and 
Jnl and Files for period covered; Ltr, Gen Heavey 
to CofEngrs USA, 30 Jun 44, sub: Rpt of Combat 
Opns, DRB AGO; MacKechnie, Notes on Nassau 
Bay-Mubo-Tambu Bay-Salamaua Opns, 29 Jun— 12 
Sep 43, no date, DRB AGO. 

of troubles. The rain obscured the guide 
lights on the offshore islands. The escort- 
ing PT lost the convoy. The lead PT 
overshot Nassau Bay. Some of the land- 
ing craft of the first wave followed it, 
then lost time turning around and find- 
ing the convoy again. 

The landing began, in rainy darkness, 
shortly after midnight. The Australian 
platoon on shore had lost its way and 
arrived at Nassau Bay, in time to install 
only two instead of three lights. Thus 
the first two waves of landing craft in- 
termingled and landed together on the 
same stretch of beach. And a ten- to 
twelve-foot surf, a rare occurrence at Nas- 
sau Bay, was pounding. It rammed the 
landing craft so far up on the beach that 
seventeen of them could not back off but 
promptly broached and filled with water, 
almost complete wrecks. The LCM, after 
unloading a bulldozer, was able to re- 
tract; it proceeded out to sea and got 
the troops off the lead PT boat, and then 
returned to the beach where it swamped. 

There was no enemy opposition, nor 
any casualties. Japanese in an outpost at 
the beach had fled into the jungle, be- 
lieving, prisoners reported later, that 
the bulldozer was a tank. Except for the 
landing craft, there were no serious losses 
of equipment, but most of the radios 
were damaged by salt water. 

Seven hundred and seventy men were 
landed that night. 20 The leader of the 
third wave, which arrived hours after 
the first two, realized that his craft were 
the only ones immediately available for 
resupply and decided not to land until 
the surf abated. He took the barge and 
the rest of the LCVP's, with B Company 

"The first report gave 740 but was soon cor- 
rected. See msgs in 41st Div G— 3 Jnl, 30 Jun 43. 



on board, to shelter in a cove down the 
coast. When the storm subsided they re- 
turned to Nassau Bay but failed to make 
contact with the troops, who were beat- 
ing off a Japanese attack. The wave re- 
turned to Mageri Point, then went back 
to Nassau Bay and landed on the after- 
noon of 2 July. 

Once on shore A and C Companies, 
i62d Infantry, established defense lines 
three hundred yards north and south, 
respectively, of the landing beach. The 
Australian platoon defended the west (in- 
land) flank. There was no contact with 
the enemy that night. By daybreak of 
30 June the beach was cleared of all 
ammunition, equipment, and supplies. 
Beach defenses, employing machine guns 
salvaged from the wrecked landing craft, 
were set up. Communication with higher 
headquarters was a problem. Most of the 
water-soaked radios would not work, and 
during the first few days Colonel Mac- 
Kechnie was out of contact with New 
Guinea Force, 41st Division headquar- 
ters, and Morobe at one time or another. 
Nothing was heard from the Papuan In- 
fantry Battalion elements on the other 
side of Cape Dinga for several days. All 
the SCR's 511 and 536, the small hand 
radios used for tactical communication 
within infantry battalions, had been 
soaked and were never usable during 
the subsequent operations against Sala- 

After daylight of go June C Company 
marched south to the Tabali River just 
west of Cape Dinga. Company A started 
north from its night positions to clear 
the area as far as the south arm of the 
Bitoi River but soon ran into enemy 
mortar and machine gun fire (its first 
such experience) and halted. Patrols went 

out and reported the enemy as present 
in some strength. Then A Company, re- 
inforced by a platoon of D Company, 
2/6th Australian Infantry Battalion of 
the 17th Brigade, which had flashed the 
landing lights, attempted to strike the 
Japanese right (west) flank but was 
stopped. When the Australian platoon 
ran out of ammunition it was relieved by 
a detachment of engineers from the crews 
of the wrecked landing craft. Two of the 
C Company platoons came up from the 
south to join A Company. At 1500 the 
force started forward and by 1650 had 
brushed away scattered Japanese oppo- 
sition to reach the south arm of the Bitoi 

When General Adachi received word 
of the invasion his first thought was to 
destroy the MacKechnie Force before it 
had a chance to consolidate. But General 
Nakano persuaded him that it would be 
better to "delay the enemy advance in 
NASSAU from a distance" and to con- 
centrate on the Australian threat at Bob- 
dubi. 21 So no more enemy troops were 
sent against MacKechnie. Meanwhile the 
Papuan Infantry Battalion troops began 
pressing against the rear of the Japanese 
detachment at Cape Dinga. This detach- 
ment began moving toward the Amer- 
ican beachhead. 

About 1630 the C Company platoon 
defending the left (south) flank reported 
that Japanese troops were crossing the 
Tabali River just south of its position, 
whereupon it was ordered to withdraw 
to the south flank of the landing beach 
proper to hold a line between the beach 
and a swamp which began a short dis- 

31 18th Army Operations, II, Japanese Monogr 
No. 4a (OCMH), 14. 



tance inland. Before the platoon could 
move, Japanese troops attacked its rear 
and flank. The platoon fought its way 
north, losing its commander and four 
enlisted men killed on the way. 

While the platoon was withdrawing, 
Capt. Paul A. Cawlfield, MacKechnie 
Force S-3, organized a defense line at the 
beach using engineers, part of D Com- 
pany, and men from force headquarters. 
At dusk the harassed platoon reached 
this line, and then the enemy struck in a 
series of attacks that lasted all night. 
Machine gun, mortar, and rifle fire and 
grenades hit the American positions, and 
small parties attempted to infiltrate. But 
the American units, in action for the 
first time, beat off the attackers who, 
except for scattered riflemen that were 
hunted down and killed, pulled out just 
before sunrise. The MacKechnie Force 
estimated that it had killed fifty Jap- 
anese. Its own casualties were eighteen 
killed, twenty-seven wounded. Colonel 
MacKechnie later asserted that in his 
opinion several of the American casual- 

ties were caused by American troops fir- 
ing at each other in the excitement of 
the night action. 

By 2 July, with the landing of B Com- 
pany and other elements of the third 
wave, the Nassau Bay beachhead was 
considered secure. On that date the 
Americans made contact with the 17th 
Brigade, and the MacKechnie Force 
made ready to execute its missions in 
the northward drive against Salamaua. 

Thus with the landings at Woodlark, 
Kiriwina, and Nassau Bay, General Mac- 
Arthur's Southwest Pacific Area inaug- 
urated Cartwheel. Compared with the 
massive strokes of 1944 and 1945, the 
operations were small, but they gave in- 
valuable amphibious experience to sol- 
diers and sailors and they began a for- 
ward movement that was not halted until 
final victory. 

Meanwhile, on the other side of the 
Solomon Sea, Admiral Halsey's South 
Pacific forces had executed their first 
Cartwheel missions by invading New 


Toenails: The Landings in New Georgia 

The South Pacific's tactical and logis- 
tical planning for the invasion of New 
Georgia (Toenails, or Operation A) in- 
volved all the major echelons of the 
complex command that was Admiral 
Halsey's. Halsey's position was somewhat 
unusual. As he phrased it, the Joint 
Chiefs' orders of 28 March "had the curi- 
ous effect of giving me two 'hats' in the 
same echelon." 1 His immediate superior 
in the chain of command was Admiral 
Nimitz, who was responsible, subject to 
decisions by the Joint Chiefs, for supply- 
ing him with the means of war. For the 
strategic direction of the war in the Solo- 
mons MacArthur was Halsey's superior. 

South Pacific Organization 

Whereas MacArthur's headquarters 
followed U.S. Army organization, Hal- 
sey's followed that of the Navy. 2 There 
were many more subordinates, such as 
island commanders, reporting directly 
to Halsey than reporting to MacArthur, 
and the South Pacific was never organ- 

1 Halsey and Bryan, Admiral Halsey's Story, p. 

1 Unless otherwise indicated this section is based 
on The History of the United States Army Forces 
in the South Pacific Area [USAFISPA] During 
World War II: 30 March 1942-1 August 1944, MS, 
Pt. II, Chs. I-II, IV, and Pt. Ill, Vol. I, Ch. I, 

ized as simply as the Southwest Pacific. 
Halsey, by the device of not appointing 
a single tactical commander of all naval 
forces, retained personal control of them. 
There was a single commander of land- 
based aircraft, but there was never a 
single ground force commander with 
complete tactical authority. \(Chart ?y| 

Naval forces, designated the Third 
Fleet in March 1943, came generally from 
the U.S. Navy and the Royal New Zea- 
land Navy. Except for New Zealand 
ships, no warships were ever perma- 
nently assigned; as need arose Nimitz 
dispatched warships to the South Pacific. 
The South Pacific Amphibious Force 
(Task Force 32), on the other hand, was 
a permanent organization to which land- 
ing forces were attached for amphibious 
operations. In command was Rear Adm. 
Richmond K. Turner who had led the 
Amphibious Force in the invasion of 
Guadalcanal the year before. 

Land-based air units from all Allied 
services in the South Pacific were under 
the operational control of the Com- 
mander, Aircraft, South Pacific, Admiral 
Fitch. Fitch's command, Task Force 33, 
was made up of Royal New Zealand and 
U.S. Army, Navy, and Marine Corps air 
units. Principal administrative organiza- 
tions within Task Force 33 were General 
Twining's Thirteenth Air Force and the 

Chart 5— Organization of Principal South Pacific Forces, June 1943 


Corps 2 

Ground troops 

Allied nations ♦ 


Joint task 
forces for operations 
amphibious phase 

Services of 
South Pacific Area 

1 No commander appointed by COMSOPAC. 

1 Administrative responsibility for Marine ground units; tactical control only far Bougainville. 
3 Administrative responsibility for all Army units, ground and air,- tactical control occasionally assigned by 

* For operations, served under I Marine Amphibious Force, XIV Corps, and islond base commanders. 



ist and 2d Marine Air Wings. The most 
important tactical organization in Fitch's 
force was the interservice, international 
outfit known as Air Command, Solo- 
mons, that had grown out of the exigen- 
cies of the Guadalcanal Campaign. 3 Fitch 
issued general directives which were 
executed under the tactical direction of 
the Commander, Aircraft, Solomons, 
who until 25 July 1943 was Rear Adm. 
Marc A. Mitscher. 

There were two principal ground force 
commanders in early 1943. The first, 
General Harmon, an experienced air- 
man who had served as Chief of Air 
Staff in Washington, was the command- 
ing general of U.S. Army Forces in the 
South Pacific Area; his command em- 
braced air as well as ground troops. His 
authority was largely administrative and 
logistical, but he also advised the area 
commander on tactical matters and Hal- 
sey throughout the period of active oper- 
ations relied heavily on him. Under Har- 
mon, in early 1943, were four infantry 
divisions, the Americal, 25th, 37th, and 
43d, as well as the Thirteenth Air Force. 
The Americal and 25th Divisions had 
fought in the Guadalcanal Campaign. 
The 43d Division had seen no fighting 
but had received valuable experience 
when elements of the division took part 
in the invasion of the Russells. The 37th, 
which had gone out the year before to 
garrison the Fijis, w T as as yet untried. In 
addition to these divisions, which usually 
fought under the tactical command of 
the XIV Corps, there were, in Harmon's 
command, the Army garrison troops in 
the island bases and a growing number, 

3 During the Guadalcanal Campaign the senior 
naval aviator on the island commanded all aircraft 
there, with the informal title of Comair Cactus. 

but never enough to satisfy the local 
commanders, of service units. By mid- 
1943 Harmon's command embraced 
about 275,000 men. 

The Marine Corps counterpart to 
Harmon's command, as far as ground 
forces were concerned, was the I Marine 
Amphibious Corps. This organization, 
under Maj. Gen. Clayton B. Vogel, 
USMC, had administrative responsibility 
over all Marine Corps units, except ships' 
detachments and certain air units, in the 
South Pacific— two Marine divisions, one 
raider regiment, six defense battalions, 
one parachute regiment, and service 
troops. The 1st Marine Division in the 
Southwest Pacific was nominally admin- 
istered by the I Marine Amphibious 
Corps but drew its supplies from South- 
west Pacific agencies. 

The highest logistic agency, the Serv- 
ice Squadron, South Pacific Force, oper- 
ated directly under Halsey. It controlled 
all ships, distributed all supplies locally 
procured, assigned shipping space, desig- 
nated ports, and handled all naval pro- 
curement. An equally important logistic 
agency was the Army's Services of Supply, 
South Pacific Area. In early 1943 under 
Maj. Gen. Robert G. Breene it was an 
expanding organization which was play- 
ing an important part in South Pacific 

The organization of the South Pacific, 
as set forth on paper, seems complicated 
and unwieldy. Perhaps it could have 
functioned awkwardly, but the personali- 
ties and abilities of the senior command- 
ers were such that they made it work. 
There is ample testimony in various re- 
ports to attest to the high regard in which 
the aggressive, forceful Halsey and his 
subordinates held one another, and 



21 June - 5 July 1943 

US Landings 


O Airfields 






Mbulo I 

MAP 7 

Printed by Army Map Service 



events showed that the South Pacific was 
able to plan and conduct offensive opera- 
tions involving units from all Allied 
armed services with skill and success. 

Preparations and Plans 

Admiral Halsey and his officers had 
begun planning and preparing for New 
Georgia in January 1943, before the end 
of the Guadalcanal Campaign. This proc- 
ess, which involved air and naval bom- 
bardments, the assembly of supplies, and 
reconnaissance of the target area, as well 
as the preparation and issuance of opera- 
tion plans and field orders, continued 
right up to D Day, 30 June. 

The Target 

In climate, topography, and develop- 
ment, the Solomons are much like New 
Guinea and the Bismarcks. Their in- 
teriors were virtually unexplored. They 
are hot, jungled, wet, swampy, moun- 
tainous, and unhealthful. 4 

New Georgia is the name for a large 
group in the central Solomons which 
includes Vella Lavella, Gizo, Kolom- 
bangara, New Georgia (the main island 
of the group), Rendova, and Vangunu, 
Simbo, Ganonnga, Wana Wana, Arun- 
del, Bangga, Mbulo, Gatukai, Tetipari 
(or M ontgomery '), and a host of islets and 
reefs. (Map y) From Vella Lavella to 
Gatukai, the cluster is 1 25 nautical miles 
in length. Several of the islands have sym- 
metrical volcanic cones rising over 3,000 
feet above sea level. 

In addition to the multitude of small 
channels, narrows, and passages, navi- 

* Data for this subsection are taken from U.S. 
Navy hydrographic charts; military maps; MIS 
"WDGS, Survey of the Solomon Islands; a terrain 
study in 37 th Div records. 

gable only by small craft, there are sev- 
eral large bodies of water in the group. 
The Slot, the channel sailed so frequently 
by the Japanese during the Guadalcanal 
Campaign, lies between New Georgia 
on one side and Choiseul and Santa Isa- 
bel on the other. Marovo Lagoon on New 
Georgia's northeast side is one of the 
largest in the world. Vella Gulf separates 
Vella Lavella from Kolombangara, which 
is set off from New Georgia by Kula Gulf. 
Blanche Channel divides New Georgia 
from Rendova and Tetipari. 

The island of New Georgia proper, 
the sixth largest in the Solomons, is about 
forty-five statute miles long on its north- 
west-southeast axis, and about thirty 
miles from southwest to northeast. It is 
mountainous in the interior, low but 
very rough in the vicinity of Munda 

New Georgia proper was difficult to 
get to by sea except in a few places. Reefs 
and a chain of barrier islands blocked 
much of the coast line, which in any event 
was frequently covered by mangrove 
swamps with tough aerial prop roots. 
The best deepwater approach was the 
Kula Gulf which boasted a few inlets, 
but Japanese warships and seacoast guns 
defended much of the shore line of the 
gulf. There were protected anchorages 
in the southeast part of the island at 
Wickham Anchorage, Viru Harbor, and 
Segi Point. Munda Point, the airfield 
site, was inaccessible to large vessels. 
East and west of the point visible islets 
and reefs, and also invisible ones, barred 
Roviana and Wana Wana lagoons to 
large ships. Rounding the lagoons like a 
crude fence on the seaward side is a 
tangled string of islands, rocks, and coral 
reefs— Roviana, Sasavele, Baraulu, and 



others, some with names, some without. 
These all have cliffs facing the sea (south) 
and slope down to sea level on the lagoon 
side. The channels between the barrier 
islands were too shallow for ships. Nor 
could ships reach Munda Point from 
Kula Gulf and Hathorn Sound. Diamond 
Narrows, running from Kula Gulf to 
the lagoons, was deep but too narrow 
for large vessels. 

Across Blanche Channel from Munda 
and her guardian islands lies mountain- 
ous Rendova, which could be reached 
from the Solomon Sea. Rendova Harbor, 
though by no means a port, offered an 
anchorage to ocean-going ships. 

During the first months of 1943 coast- 
watchers covered the Solomons thor- 
oughly. Buka Passage, between Bougain- 
ville and Buka, and Buin on southern 
Bougainville had been the sites of coast- 
watching stations for several months, and 
in October 1942 flying boats and sub- 
marines took watchers to Vella Lavella, 
Choiseul, and Santa Isabel. 5 

At Segi Point on New Georgia was 
Donald G. Kennedy, a New Zealander 
who was District Officer in the Protec- 
torate Government. Like Resident Com- 
missioner William S. Marchant, the An- 
glican Bishop of Melanesia, and various 
other officials and members of religious 
orders, Kennedy remained in the Solo- 
mons when the Japanese came. 6 At Segi 
Point Kennedy organized a network of 
white and Melanesian watchers cover- 

3 By July, unfortunately, the Japanese were hunt- 
ing the Bougainville coastwatchers so resolutely 
that the stations there had to be abandoned. See 
Feldt, The Coastwatchers, Ch. XI. 

" [British] Central Office of Information, Among 
Those Present: The Official Story of the Pacific 
Islands at War (London: His Majesty's Stationery 
Office, 1946), pp. 11, 43. 

ing Kolombangara, Rendova, Vangunu, 
Santa Isabel, and Roviana. A Euronesian 
medical practitioner was posted on Santa 
Isabel. On Roviana Sgt. Harry Wickham 
of the British Solomon Islands Defense 
Force organized the natives to keep watch 
over Munda Point. 

Kennedy raised a guerrilla band to 
protect his hideout at Segi Point, for 
the Japanese occasionally sent out puni- 
tive expeditions to hunt him down. The 
primary mission of the coastwatchers 
was watching, not fighting, but Kennedy 
and his band were strong enough to 
wipe out several patrols that came too 
close. On one occasion Kennedy and his 
men, aboard the ten-ton schooner Dada- 
vata, saw a Japanese whaleboat systemat- 
ically reconnoitering the islets in Marovo 
Lagoon. They attacked \vith rifles, 
rammed the whaleboat, sank it, and 
killed or drowned its company. 7 

In addition to gaining information 
from terrain studies, interrogation of 
former residents, and coastwatchers' re- 
ports, South Pacific headquarters was 
able to augment its knowledge of New 
Georgia by a series of ground patrols. 
The first such expedition was directed 
by General Vogel. Four officers and eight 
enlisted men from each of the four bat- 
talions of the 1st Marine Raider Regi- 
ment assembled on Guadalcanal on 17 
March, then sailed to Florida to board 
amphibian patrol planes (PBY's) which 
took them to Segi Point. After Kennedy 
furnished them with native scouts and 
bearers, patrols went out to reconnoiter 
Kolombangara, Viru Harbor, Munda 
Point, and other areas. Traveling over- 
land and by canoe, they carefully exam- 

T Ibid., p. 52. 



ined caves, anchorages, and passages. 
Their mission completed, all parties re- 
assembled at Segi Point on g April. 

The raiders' reports indicated that 
troops in small craft could be taken 
through Onaiavisi Entrance to a 200- 
yard-long beach at Zanana, east of the 
Barike River. From there they could 
strike westward toward Munda. 8 Before 
D Day, additional patrols from the invad- 
ing forces went to New Georgia and 

From November 1942 until D Day, 
Munda and Vila airfields were continu- 
ously subjected to air and naval bom- 
bardments. Vila, located in a swampy 
region, was practically never used by 
the enemy. From January until D Day, 
Allied cruisers and destroyers shelled 
Munda four times at night, Vila three 
times. The net result of the continuous 
air bombardment and the sporadic naval 
shelling was that the Japanese could not 
base planes permanently at Munda. It 
was used, and only occasionally, as a for- 
ward staging field. 9 

Logistic Preparations 

On Halsey's orders South Pacific agen- 
cies had begun assembling supplies and 
developing bases and anchorages for the 
invasion of New Georgia as early as Jan- 

8 1 Mar Amphib Corps, Report on New Georgia 
Ground Reconnaissance: 21 March-g April 43, 18 
Apr 43. 

9 Morison, The Struggle for Guadalcanal, pp. 
33^-47; Morison, Breaking the Bismarchs Barrier, 
pp. 106-10; U.S. Strategic Bombing Survey, Inter- 
rogations of Japanese Officials (Washington, 1946), 
I, 142, 192; USSBS, The Allied Campaign Against 
Rabaul, p. 43; USSBS, The Thirteenth Air Force in 
the War Against Japan, p. 6; Southeast Area Naval 
Operations, II, Japanese Monogr No. 49 (OCMH), 
19; Southeast Area Air Operations, 1942-44, Jap- 
anese Monogr No. 38 (OCMH), pp. 7—11. 

uary 1943. Admiral Turner, remember- 
ing his experiences in the Guadalcanal 
Campaign, suggested that supplies for 
the invasion be stockpiled on Guadal- 
canal, and in February movement of 
supplies to Guadalcanal (under the ap- 
propriate code name Drygoods) began. 
In spite of the fact that the port of Nou- 
mea, New Caledonia, was jammed with 
ships waiting to be unloaded, in spite of 
the fact that port facilities at Guadal- 
canal were so poor, and in spite of a 
bad storm at Guadalcanal in May that 
destroyed all the floating quays, washed 
out bridges, and created general havoc, 
enough supplies for the invasion were 
ready on Guadalcanal by June. This 
was accomplished by Herculean labor at 
Noumea, by routing some ships directly 
to Guadalcanal, and by selective dis- 
charge of cargo from other ships. The 
effects of the storm at Guadalcanal were 
alleviated by using the ungainly-look- 
ing 2I/2 -ton, six-wheel amphibian truck 
(DUKW) to haul supplies from ships to 
inland dumps over open beaches. By 
June 54,274 tons of supplies, exclusive 
of organization equipment, maintenance 
supplies, and petroleum products dis- 
charged from tankers, had been put 
ashore. In addition many loaded vehicles, 
13,085 tons of assorted gear, and 23,775 
drums of fuel and lubricants were moved 
from Guadalcanal, to the Russells in 
June. Bulk gasoline storage tanks with 
a capacity of nearly 80,000 barrels were 
available on Guadalcanal. 10 Although 

"COMSOPAC War Diary, 1 Jan 43 entry; Ex- 
tract of recommendations submitted by COMAM- 
PHIBFORSOPAC, Incl F to memo, Gen Peck for 
Gen Breene et al., 16 Jan 43, sub: Notes on Mtg 
Held in War Plans See COMSOPAC, 14 Jan 43, in 
USAFISPA File No. 381, Preliminary Ping COM- 



Noumea and Espiritu Santo in the New 
Hebrides were still the main South Pa- 
cific bases, Guadalcanal was ready to play 
an important role. The South Pacific 
commanders had insured that haphazard 
supply methods would not characterize 

T actical Plans 

Final plans and orders for Toenails 
were ready in June. 11 Halsey had hoped 
to invade New Georgia in April, but 
could not move before the Southwest 
Pacific was ready to move into the Tro- 
briands and Nassau Bay. The general 
concept of the operation was worked out 
by Admiral Halsey, a planning commit- 
tee, and members of Halsey's staff. The 
committee consisted of General Harmon, 
the Army commander; Admiral Fitch, 
the land-based air commander; Admiral 
Turner, the amphibious commander; 
and General Vogel of the I Marine Am- 
phibious Corps. The principal staff offi- 
cers concerned were Admiral Wilkinson; 
Captain Browning, Halsey's chief of staff; 

Ltr, COMSOPAC to COMGENSOPAC et al., 24 
Feb 43, sub: Assembly of Sups for Future Opns, in 
USAFISPA G-2 Hist Sec File, Ping for New 
Georgia Opn, OCMH; The History of USAFISPA, 
Pt. I, Vol. I, p. 178, and Pt. Ill, pp. 649-51, 661, 669, 
673-74, OCMH; ONI USN, Operations in the New 
Georgia Area, p. 3. 

" Two flies previously cited, USAFISPA G-2 Hist 
Sec File, Ping for New Georgia Opn, and USAFI- 
SPA File No. 381, Preliminary Ping COMSOPAC 
and COMGENSOPAC, contain valuable material for 
the student interested in the genesis and develop- 
ment of tactical plans. This subsection is based on 
these two files and on Hq NGOF FO 1, 16 Jun 43, 
and Addendum 1, 24 Jun 43; 43d Div FO 1, 17 
Jun 43; COMSOPAC Opn Plan 14-43, 3 Jun 43, 
with annexes; CTF 31 Opn Plan A8-43, 4 Jun 43, 
with annexes; CTG 31.3 Opn Order AL 10-43, 21 
Jun 43; CTG 31.1 Loading Order 1—43, 13 Jun 43; 
CTG 31.3 Loading Order 1—43, 16 Jun 43. Last five 
in Off of Naval Reds and Library, 

and General Peck, Halsey's war plans 
officer. By May agreement was reached 
on the general plan. It called for the 
simultaneous seizure of Rendova, Vim 
Harbor, Wickham Anchorage, and Segi 
Point. A fighter field would be built at 
Segi Point. After the initial landings 
small craft from Guadalcanal and the 
Russells would stage through Wickham 
Anchorage and Viru Harbor to build 
up Rendova's garrison. Munda's field 
would be harassed and neutralized by 
155-mm. guns and 105-mm. howitzers 
emplaced on Rendova and the nearer 
barrier islands. These moves were prep- 
aratory to the full-scale assaults against 
Munda and Vila, and later against south- 
ern Bougainville. 

Assigned to the operation were South 
Pacific aircraft, warships, the South Pa- 
cific Amphibious Force, and the heavily 
reinforced 43d Division with its com- 
mander, Maj. Gen. John H. Hester, in 
command of the landing forces. The 37th 
Division, less elements, was in area re- 
serve to be committed only on Halsey's 

Final plans and tactical organization 
were complicated, as Toenails called for 
four separate simultaneous invasions 
(Rendova, Wickham Anchorage, Segi 
Point, and Viru Harbor) with the Ren- 
dova landing to be followed by two more 
on the same island. 

Admiral Halsey's basic plan, issued on 
3 June, organized the task forces, pre- 
scribed their general missions, and di- 
rected Admiral Turner to co-ordinate the 
planning of the participating forces. Four 
task forces w T ere assigned to the opera- 
tion: Task Force 33, the Aircraft, South 
Pacific, under Admiral Fitch; Task Force 
72, a group of Seventh Fleet submarines 


Chart 6 — Organization of South Pacific Forces for Toenails 


TF 31 
Attack Force 
Adm Turner 

5 task groups 

TG 36.6 
Gen Force 
Gen Beightler 





9 Cruisers 
29 DD's 
2 Tankers 

37th Div 
-129th RCT 
-148th RCT (-1 
-136th FA Bn 


commanded by Capt. James F. Fife and 
now under Halsey's operational control; 
Task Force 36, the naval covering force 
commanded, in effect, by Halsey himself; 
and Task Force 31, the attack force. 
(Chart 6) 

Task Force 33, to which Halsey tem- 
porarily assigned planes from Carrier 
Division 22 (three escort carriers), was 
to provide defensive reconnaissance for 
New Georgia operations and the South- 
west Pacific's seizure of Woodlark and 
Kiriwina, and to cover the area north- 
east of the Solomons (Southwest Pacific 
planes were responsible for the Bis- 
marcks). It was to destroy enemy units 
which threatened South and Southwest 

Pacific forces, especially Japanese planes 
operating from New Georgia and south- 
ern Bougainville. Fitch's planes were also 
to provide fighter cover, direct air sup- 
port, and liaison and spotting planes for 
the attack force. Starting D minus 5, Task 
Force 33 would attempt to isolate the 
battlefield by attacking the Japanese air 
bases at Munda, Ballale, Kahili, Kieta, 
and Vila, and by striking at surface ves- 
sels in the Bougainville and Munda 
areas. During daylight, fighters would 
cover ships and ground troops, and anti- 
submarine patrols would be maintained 
for convoys. Black Cats (PBY's) would 
cover all night movements. Striking 
forces at all times were to be prepared 



to hit enemy surface ships. Beginning 
on D Day, eighteen dive bombers would 
remain on stand-by alert in the Russells. 
Medium bombers were to be prepared 
to support the ground troops. Finally, 
arrangements were made for air drop- 
ping supplies and equipment to the 
ground troops in New Georgia. 

One innovation in the command of 
supporting planes had apparently arisen 
from Maj. Gen. Alexander A. Vande- 
grift's recommendations based on his 
experiences in invading Guadalcanal. 12 
Halsey directed that on take-off from 
Guadalcanal and Russells fields planes 
assigned to missions in the immediate 
area of operations would come under 
control of the local air commander (the 
Commander, New Georgia Air Force). 
Direction of fighters over Task Force 31 
was to be conducted by a group aboard 
a destroyer until direction could be con- 
ducted ashore on Rendova. Similarly, 
bomber direction for direct air support 
would be handled aboard Turner's flag- 
ship McCawley until bomber director 
groups could establish themselves ashore. 
In early June, Fitch issued orders con- 
centrating most of his strength in the 
Guadalcanal area under Admiral Mit- 
scher. 13 Totals for aircraft involved were 
fairly impressive. On 30 June Fitch had 
on hand for the operation 533 planes, of 
which 213 fighters, 170 light bombers, 

" See Jeter A. Isely and Philip A. Crowl, The U.S. 
Marines and Amphibious War: Its Theory, and Its 
Practice in the Pacific (Princeton, N.J.: Princeton 
University Press, 1951), p. 172, and Miller, Guadal- 
canal: The First Offensive, p. 71. 

43, sub: Availability of Aircraft at Mainyard 
[Guadalcanal] for Toenails Opn, in USAFISPA 
G— 2 Hist Sec File, Ping for New Georgia Opn, 

and 72 heavy bombers were ready to 
fly. 14 

Task Force 36 included part of the 
37th Division on Guadalcanal in area 
reserve, besides all Halsey's naval 
strength except that assigned to the at- 
tack force. Naval units, including air- 
craft carriers (two CV's and three 
CVE's), battleships, cruisers, and de- 
stroyers, would operate out of Noumea, 
New Caledonia, and the New Hebrides 
into the Coral and Solomon Seas to inter- 
cept and destroy any Japanese forces 
which ventured out. The reserve 37th 
Division forces were to be committed, 
on five days' notice, on orders from 

Captain Fife's submarines would at 
first conduct offensive reconnaissance 
from about latitude one degree north 
southward to the prevailing equatorial 
weather front. Once the Japanese were 
aware of the invasions, Fife's boats were 
either to concentrate on locating enemy 
vessels or to withdraw south to cover 
Bougainville Strait and the waters be- 
tween New Ireland and Buka. This re- 
connaissance would be in addition to 
patrols by Central Pacific submarines, 
which would keep watch over any Jap- 
anese surface forces approaching the 
South from the Central Pacific. 

Admiral Turner's attack force (Task 
Force 31) consisted of ships and landing 
craft from the South Pacific or III Am- 
phibious Force (Task Force 32), plus 
the ground troops. These troops, desig- 
nated the New Georgia Occupation 
Force, initially included the following 

" ONI USN, Operations in the New Georgia Area, 
p. 62. 



43d Division 

gth Marine Defense Battalion 

1st Marine Raider Regiment (less two 

136th Field Artillery Battalion (155- 

mm. howitzers), 37th Division 
Elements of the 70th Coast Artillery 

Battalion (Antiaircraft) 
One and one-half naval construction 

Elements of the 1st Commando, Fiji 

Guerrillas 15 
Radar units 

Naval base detachments 
A boat pool 

Creating the New Georgia Occupa- 
tion Force, and attaching all ground 
troops to it (instead of attaching the sup- 
porting units to the 43d Division), made 
another headquarters necessary, and 
threw a heavy burden on 43d Division 
headquarters. General Hester com- 
manded both force and division, and the 
43d Division staff was, in effect, split into 
two staffs. The 43d Division's staff sec- 
tion chiefs (the Assistant Chiefs of Staff, 
G-i, G-2, G-3, and G-4), as well as offi- 
cers from Harmon's headquarters, served 
on the Occupation Force staff sections, 
and their assistants directed the division's 
staff sections. Brig. Gen. Harold R. Bar- 
ker, 43d Division artillery commander, 
commanded all Occupation Force artil- 
lery—field, seacoast, and antiaircraft. 

From the start General Harmon was 

" This unit included besides the Fijians some 
Tongans and a few Solomon Islanders. See [British] 
Central Office of Information, Among Those Pres- 
ent, pp. 53—56. American documents list this unit 
variously as "ist South Seas Scout Company," "South 
Pacific Scouts," "native troops," and erroneously, as 
"1st Company, ist Fiji Infantry," which was a dif- 
ferent unit serving at Port Purvis on Florida. 

dubious about the effectiveness of this 
arrangement. He was "somewhat con- 
cerned that Hester did not have enough 
command and staff to properly conduct 
his operation in its augmented con- 
cept." 16 On 10 June, with Halsey's con- 
currence, he therefore told Maj. Gen. 
Oscar W. Griswold, commanding the 
XIV Corps and the Guadalcanal Island 
Base, to keep himself informed regard- 
ing Hester's plans in order to be pre- 
pared to take over if need be. 17 

The general plan of maneuver called 
for assault troops from Guadalcanal and 
the Russells to move to Rendova, Segi 
Point, Wickham Anchorage, and Viru 
Harbor on APD's, transports, cargo ships, 
minesweepers, and minelayers. Segi, 
Wickham, and Viru would be taken by 
small forces to secure the line of commu- 
nications to Rendova while the main 
body of ground forces captured Rendova. 
Artillery on Rendova and the barrier 
islands was to bombard Munda, an ac- 
tivity in which ships' gunfire would also 
be employed. On several days following 
D Day, slow vessels such as LST*s and 
LCT's would bring in more troops and 
supplies. They would travel at night and 
in daylight hours hide away, protected 
from Japanese planes by shore-based anti- 
aircraft, in Wickham Anchorage and 
Viru Harbor. About D plus 4, when 
enough men and supplies would be on 
hand, landing craft were to ferry assault 
troops from Rendova across Roviana La- 
goon to New Georgia to begin the march 

18 Ltr, Harmon to Handy, 15 Jul 43, quoted in 
part in Hq SOPACBACOM File, Suppl New Georgia 
Material, OCMH. 

" Rad, COMGENSOPAC to COMGEN Mainyard, 
10 Jun 43, Hq SOPACBACOM File, Suppl New 
Georgia Material, OCMH. 


Chart 7 — Organization of Attack Force, D Day 1 



TF 31 
Force (Turner) 

TG 31.2 

(4 DD's) 

TG 31,1 


4 DD's 


.ter) 2 

NGOF Reserve 

TG 31.3 

TG 31.4 
Reserve MTB 

1st Mar Rdr Regt (-) 

1 This chart shows D-Day organization only. Second and third echelons were organized along similar lines. 

2 To assume control of all assigned land, sea, and air forces in New Georgia on orders from Admiral Halsey. 

against Munda. Coupled with this ad- 
vance would be the amphibious seizure 
of Enogai Inlet in the Kula Gulf to cut 
the Japanese reinforcement, supply, and 
evacuation trail between Munda and 
Enogai, and thus prevent the Japanese 
on Kolombangara from strengthening 
their compatriots on New Georgia. Once 
Munda and Enogai were secured, it was 
planned, Vila on Kolombangara would 
be seized and further advances up the 
Solomons chain would follow. 

Turner organized his force into five 
groups. (Chart 7) The Western Force 
(Task. Group 31.1), which Turner com- 

manded in person, would seize Rendova 
and make subsequent assaults against 
Munda, Enogai, and Kolombangara. The 
Eastern Force, under Rear Adm. George 
H. Fort, was to take Segi, Viru, and 
Wickham. Task Group 31.2, consisting 
of eight destroyers, would cover the trans- 
ports. No ships' gunfire support was 
planned in advance, but all ships, in- 
cluding transports, were ordered to be 
ready to deliver supporting and counter- 
battery fire if necessary. 

The New Georgia Occupation Force, 
under General Hester, included the 
Western Landing Force (under Hester), 



which during the amphibious phase 
would function as part of Turner's West- 
ern Force; the Eastern Landing Force 
(under Col. Daniel H. Hundley), which 
during the amphibious phase would be 
part of Fort's Eastern Force; naval base 
forces for all points to be captured; the 
reserve under Col. Harry B. Liversedge, 
USMC; and two more whose designa- 
tions are not self-explanatory— the New 
Georgia Air Force and the Assault Flo- 

tillas. (C harts] 8\ and 9) 

The New Georgia^ir Force, led by 
Brig. Gen. Francis P. Mulcahy, USMC, 
consisted initially of Headquarters, 2d 
Marine Air Wing. In contrast with the 
system in the Southwest Pacific, this air 
headquarters was under the landing force 
commander. Mulcahy was to take over 
control of New Georgia air operations 
during the amphibious phase once that 
control was relinquished by Turner; he 
would take command of the planes from 
Guadalcanal and the Russells that would 
be supporting the attack, once they were 
airborne. He was eventually to command 
the air squadrons to be based at Munda 
and Segi Point. The Assault Flotillas 
consisted of landing craft to be used to 
ferry the assault troops from Rendova 
to New Georgia proper when the attack 
against Munda was ready to begin. 

Two ground force units which Turner 
retained temporarily under his direct 
control were small forces designated to 
make covering landings. The Onaiavisi 
Occupation Unit, composed of A and 
B Companies, 169th Infantry, was to 
land from two APD's and one mine- 
sweeper on Sasavele and Baraulu Islands 
on either side of Onaiavisi Entrance to 
hold it until the day of the assault against 
the mainland through the entrance. The 

landing of the occupation unit was sched- 
uled for 0330, 30 June. The Rendova 
Advance Unit, C and G Companies (each 
less one rifle platoon), was to land from 
two APD's on Rendova at 0540 to cover 
the landing of the main body of the 
Western Landing Force. The latter, 
about 6,300 strong, was to start landing 
on Rendova at 0640, 30 June. 

Command over all air, sea, and ground 
forces in New Georgia would pass from 
Turner to Hester on orders from Halsey. 

The presence of the Drygoods stock- 
piles on Guadalcanal greatly simplified 
logistical problems. Three Army units of 
fire and thirty days' supplies were to be 
put ashore at Rendova, and five units of 
fire and thirty days' supplies at Viru, 
Segi, and Wickham. Supply levels were 
to be built to a sixty-day level out of the 
Drygoods stocks. General Griswold was 
told to make the necessary quantities 
available to Turner. Turner was respon- 
sible for the actual movement of supplies 
to New Georgia. 

Directions for unloading during the 
assault phase were simple and clear. 
Turner instructed all vessels to be ready 
for quick unloading. All ships were to 
square away before reaching the trans- 
port areas offshore, and if possible to 
work all hatches from both sides. Un- 
loading parties included 150 men for 
each cargo ship and transport, 150 men 
per LST, 50 men per LCT, and 25 men 
per LCI. The shore party totaled 300 
men. Once ashore, cargo was to be moved 
off the beaches and into inland dumps as 
fast as possible. 

Secondary Landings 

With tactical plans for Toenails 
largely ready by mid- June, the invasion 

Chart 8 — Western Force on D Day 

TG 31.1 



New Georgia 
MTB Squadron 
12 MTB's 




Landing Force 

1 APD 
1 DMS 
A and B Cos. 
169th RCT 

2 APD's 

C and G Cos (— ) 

172d Inf 

4 APA's 
2 AKA's 

Hq New Georgia 
Air Force 1 

New Georgia 
Naval Base 

169fhRCT<-) Hq, NGOF Hq, 2d Mar 

3d Bn Combat Fwd ech, 43d Div Hq Air u/mo 

Team, 103d RCT Fwd ech, 43d Div Arty 3 

172d RCT(-) 

43d Sig Co 

C Co, 118th Med Bn 

43d Ren Troop ( — ) 

9th Mar Defense Bn (-) 

To include naval forces at Segi Point, Wiekham Anchorage, and Viru Harbor after establishment ashore. 

24th Cons Bn (— ) 
Boat Pool 
Naval base units 

18 LCI's plus 
smaller craft 

Chart 9— Eastern Force on D Day 

TG 31.3 


Eastern Land' 
ins Force 

Segi Group 

(21 June) 

2 APD's 
Det Bn Hq, 
O and P Cos. 
4th Mar Raider Bn 


3 APD's; 7 LCI's.- 1 APC 
1/2 Hq Co, 4th Mar Raider Bn 
N and O Cos, 4th Mar Raider Bn 
2d Bn, Combat Team, 103d RCT 
B Btry, 1 52d FA 
Det, 70th CA (AA) 
Det, 20th Cons Bn 

Viru Group 

1 DMS; 3 APD's ( Zone [DMS] "of there— on reef) 

O and P Cos, 4th Mar Raider Bn 

B Co, 103d Inf 

1/2 D Co, 103d Inf 

Det, E Btry, 70th CA (AA) 

Det, 20th Cons Bn 



forces spent the rest of the month mak- 
ing final preparations— checking weap- 
ons and supplies, conducting rehearsals 
in the New Hebrides, and studying or- 
ders, maps, and photographs. South Pa- 
cific aircraft pounded Vila, Munda, and 
the Shortlands-Bougainville bases while 
Southwest Pacific planes continued their 
long-range strikes against Rabaul. 

Segi Point 

In the midst ot" these preparations, Ad- 
miral Turner received disquieting news 
about Segi Point, which was scheduled 
to furnish the Allies with an airfield. 
Coastwatcher Donald Kennedy reported 
on 20 June that the Japanese were mov- 
ing against his hideout and that he was 
heading for the hills. He requested 
help. 18 

Kennedy's report was correct. In early 
June a small Japanese force had gone to 
the southeast part of Vangunu to deal 
summarily with disaffected natives, and 
on 17 June half the ist Battalion, 229th 
Infantry, under a Major Nagahara or 

18 The remainder of this chapter is based on 
Haisey and Bryan, Admiral Halsey's Story, pp. 158- 
61; Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 221-29; Morison, Breaking the Bis- 
marcks Barrier, pp. 138-60; ONI USN, Operations 
in the New Georgia Area, pp. 9-23; History of the 
New Georgia Campaign, MS, 13 Aug 45, prepared 
by Hist Sec,G-2 SOPACBACOM, Vol. I, Chs. II— III, 
V, OCMH; after action rpts, jnls, int rpts, and 
hists of USAFISPA, NGOF, COMAIR New Georgia 
(ad Mar Air Wing), XIV Corps, I Mar Amphib 
Corps, 43d and 37th Divs, and subordinate units, 
and 1st Mar Raider Regt; COMSOPAC, TF 31, and 
TF 33 War Diaries; 17th Army Operations, Vol. II, 
Japanese Monogr No. 40 (OCMH); 17th Army 
Operations, Map Supplement, Japanese Monogr No. 
40A (OCMH); Southeast Area Naval Operations, 
Vol. II, Japanese Monogr No. 49 (OCMH); Out- 
line of Southeast Area Naval Air Operations, Pt. 
IV, Japanese Monogr No. 108 (OCMH). 

Hara had moved from Viru Harbor 
southeast toward Segi Point. 

As loss of Segi Point prior to D Day 
would deprive the Allies of a potential 
air base, Turner, a man of fiery energy 
and quick decision, abruptly changed 
his plans. He had originally intended to 
land the heavily reinforced 1st Battalion, 
103d Infantry, at Segi on go June to 
build the fighter field and establish a 
small naval base. But on receipt of Ken- 
nedy's call for aid, he hurriedly dis- 
patched the handiest force available, the 
4th Marine Raider Battalion (less N and 
Q Companies), from Guadalcanal in the 
fast destroyer-transports Dent and Waters 
to seize Segi and hold it. Ships and ma- 
rines wasted no time. By 2030 of the 
same day— 20 June— the ships were loaded 
and under way. Before dawn next morn- 
ing they had safely worked their way 
through Panga Bay, though both ves- 
sels scraped bottom in the reef- and rock- 
filled waters. Kennedy, still safe, had lit 
bonfires on the beach and when the ma- 
rines started ashore at 0550 he was there 
to meet them. There were no Japanese. 
The major and his men were still in the 
vicinity of Lambeti Village. 

Next morning the APD's Schley and 
Crosby brought A and D Companies of 
the 103d Infantry and an airfield survey 
section to Segi Point. Though alerted 
several times against enemy attack, the 
Segi garrison was undisturbed until 30 
June, when a series of Japanese air at- 
tacks made things lively. Construction 
of the airfield began on 30 June. Using 
bulldozers and power shovels, and work- 
ing under floodlights at night, the Sea- 
bees of the 20th Naval Construction 
Battalion had the strip ready for limited 



Airfield at Segi P 

operations as a fighter staging field by 
1 1 July. 19 

Wickham Anchorage 

The force selected for the seizure o£ 
Wickham Anchorage by Vangunu Is- 
land was ready to sail from the Russells 
on 29 June. Commanded by Lt. Col. 
Lester E. Brown, the force included Col- 
onel Brown's 2d Battalion, 103d Infan- 
try, reinforced, and N and P Companies, 
plus a headquarters detachment, of the 
4th Marine Raider Battalion. Under 
Admiral Fort aboard the Trever, the 
convoy consisted of the destroyer-trans- 
ports Schley and McKean, carrying raa- 

M "The lighting of the airstrip at night was a 
carefully figured risk. We calculated we could get 
sufficient warning from our radar to turn off the 
lights before the attackers arrived." Ltr, Col Hund- 
ley to Gen Smith, Chief of Mil Hist, 19 Oct 53, no 
sub, OCMH. 

int, New Georgia. 

rines, and seven LCI's which bore sol- 
diers. The ships cast off shortly after 1 800 
and set course for Oleana Bay, about two 
and one-half miles west by south from 
Vura village. 

Allied scouting parties had reported 
that the main Japanese concentration at 
Wickham Anchorage— one platoon of 
the 229th Infantry and a company of the 
Kure 6th Special Naval Landing Force — 
was near Vura, and had also reported 
that on the east shore of Oleana Bay a 
500-yard-long strip of solid sand offered 
a good landing beach. It had therefore 
been decided to land the troops at Oleana 
Bay and then march overland and out- 
flank the enemy positions from the west. 
There were two trails from Oleana Bay 
to Wickham Anchorage. One, which fol- 
lowed the shore line, was believed used 
by the Japanese, but a shorter one had 
been cut farther inland in April by Ken- 



nedy's men in order to get scouts into 
the Vura area. This trail was thought to 
be unknown to the Japanese, and troops 
following native guides could be ex- 
pected to cover it in five or six hours. 

Visibility was practically nonexistent 
for the Wickham-bound convoy on the 
night of 29-30 June. Rain, lashed by a 
stiff wind, fell throughout the night, 
and continued as the vessels threaded 
their cautious way through the shoals 
and reefs into Oleana Bay. At 0335, 30 
June, the ships hove to. Shortly after- 
ward the first wave of marines began to 
debark from the destroyer-transports 
into LCVP's, a task complicated by dark- 
ness, rain, high wind, and heavy seas. 
Two LCVP's were almost loaded when 
the APD commanders discovered they 
were lying off the west rather than the 
east shore of the bay. The marines re- 
boarded the destroyer-transports which 
then moved a thousand yards eastward. 

Again the marines loaded into LCVP's 
and started for the beach, which was 
obscured by rain and mist. Beach flares 
which had been set by members of the 
scouting party were invisible. Only the 
noise of the breakers indicated the direc- 
tion of the shore. But things got worse. 
As the first wave of LCVP's blindly made 
their way shoreward, the LCI's broke 
into the formation and scattered it. Un- 
able to re-form, or even to see anything, 
the LCVP coxswains proceeded on their 
own. The result was exactly what might 
be expected from a night landing in bad 
weather. The assault wave of marines 
landed in impressive disorganization. 
Six LCVP's smashed up in the heavy surf 
that boiled over coral reefs. Fortunately, 
the Japanese were not present to oppose 
the landing. There were no casualties. 

The LCI's, landing in daylight, found 
the proper beach, and by 0720 the Army 
troops were ashore. More marines had 
begun landing at 0630 at the correct 
beach. With all landing operations con- 
cluded by iooo, the ships departed. 

Three officers of the reconnaissance 
party had met the landing force and in- 
formed Colonel Brown that the Japanese 
main strength was at Kaeruka rather than 
Vura. Once the scattered troops had been 
collected, the overland advance began 
with a small column moving toward Vura 
along the coastal trail while the main 
column marched against Kaeruka over 
Kennedy's trail. The marines and sol- 
diers first met the enemy in early after- 
noon. Then ensued four days of fighting 
in the sodden jungles, with the Ameri- 
cans receiving support from dive bomb- 
ers and warships, from their own heavy 
weapons, and from the 105-mm. howitz- 
ers of the 15 2d Field Artillery Battalion 
on the beach at Oleana Bay. By the end 
of 3 July the Americans, having blasted 
the Japanese out of their entrenchments, 
were in complete possession of Wickham 
Anchorage. Many of the Japanese gar- 
rison had been killed; some escaped by 
barge, canoe, or on foot. In the seizure 
of this future staging point for landing 
craft, the marines lost twelve killed, 
twenty-one wounded. Army casualties 
are not listed. 

Viru Harbor 

When the Viru Occupation Force, the 
reinforced B Company, 103d Infantry, 
on board three destroyer-transports, 
sailed into Viru Harbor before daylight 
on 30 June, lookouts vainly scanned the 
shore line for a white parachute flare. 
This was to have signaled that the ma- 



Men of 152D Field Artillery Battalion firing a ioymm. howitzer in support 
of Colonel Brown's 2d Battalion, 103d Infantry. 

rine raider companies that landed at Segi 
Point had moved against Viru from in- 
land and seized positions flanking the 
harbor, for it had been agreed that at- 
tempting to land the infantry in frontal 
assault against the high cliffs surrounding 
the harbor would be too risky. But Lt. 
Col. Michael S. Currin, commanding the 
4th Marine Raider Battalion, had warned 
that his overland march was going slowly 
and that he might not arrive and take 
the harbor by 30 June. Thus the de- 
stroyer-transports waited just outside the 
harbor, beyond range of a Japanese shore 
battery (Major Hara had left part of 
his battalion at Viru) and at noon went 
to Segi Point where with Turner's ap- 
proval the troops went ashore. The at- 

tack force commander agreed that in 
view of the delay B Company should 
follow the marine raiders in their over- 
land march. 

Currin's men had begun the first leg 
of their twelve-mile advance from Segi 
Point to Viru Harbor in rubber boats 
on 27 June. They landed near Lambeti 
Plantation that night, and the next morn- 
ing set out on their overland march. 
Skirmishes with the Japanese, coupled 
with the difficulty of walking through the 
jungle, slowed them down. They forded 
streams, knee-deep in mud and shoulder- 
high in water. The leading elements of 
the column churned the trail into slip- 
pery ooze, so that the rear elements 
floundered and stumbled along. Thus 



it was evening of 30 June before the 
marines reached Viru Harbor, which 
they took handily the next day by a 
double envelopment supported by dive 
bombers that knocked out the Japanese 
shore battery. 20 On 4 July B Company, 
103d, which had come up from Segi 
Point, took over the defenses of Viru 
Harbor from the marines. 

Thus were the operations of the East- 
ern Force conducted, separately from 
each other and separately from those of 
the Western Force, but under Admiral 
Turner's general supervision in his ca- 
pacity of attack force commander. They 
had provided one airfield and two stag- 
ing bases. While important, they were 
undertaken only to support the seizure 
of Rendova by a substantial force, which 
was then to assault Munda and Vila. 


Admiral Turner's ships that were as- 
signed to Rendova arrived off Guadal- 
canal in the morning of 2g June. 21 They 
had come up from Efate bearing the 
assault troops of the Western Landing 
Force's first echelon. They weighed an- 
chor late that afternoon and made an 
uneventful journey through the mist and 
rain to Blanche Channel between Ren- 
dova and New Georgia. 

No enemy warships were there to op- 
pose them. Their absence had been en- 
sured by a group of cruisers, destroyers, 
and minelayers from Halsey's Task Force 

!0 According to Colonel Hundley, some naval ves- 
sels, apparently unaware of the postponement, 
sailed into Viru Harbor just as the bombing began. 
The Japanese manned the shore defenses, which 
were promptly knocked out, and were completely 
surprised by the marine attack. Ibid. 

" There were six transports, two destroyer-trans- 
ports, and eight destroyers. 

36 under Rear Adm. Aaron Stanton Mer- 
rill. Merrill's ships, on the night of 29-30 
June, had bombarded Munda and Vila, 
then ventured northwest to the Short- 
lands to shell enemy bases and lay mines. 
This action inflicted damage to the Jap- 
anese while placing a surface force in 
position to cover Turner's landings. The 
bad weather canceled the air strikes 
against the Bougainville-Shortland bases, 
but Allied planes— dive and torpedo 
bombers— were able to hit Munda and 
Vila on 30 June. 

The night of 29-30 June was short 
for the six-thousand-odd troops aboard 
Turner's ships. Reveille sounded at 
0200, more than four hours before the 
ships hove to off Renard Entrance, the 
channel leading to Rendova Harbor. 

First landings were made by the Onaia- 
visi Occupation Unit— A and B Compa- 
nies, 169th Infantry. These had come 
from the Russells in the destroyer-trans- 
port Ralph Talbot and the minesweeper 
Zane to land on Sasavele and Baraulu 
Islands before daylight in order to hold 
Onaiavisi Entrance against the day that 
the New Georgia Occupation Force made 
its water-borne movement against the 
mainland. Later in the morning B Com- 
pany's 2d Platoon outposted Roviana 
Island and the next day wiped out a 
Japanese lookout station. These landings 
were not opposed. The Japanese had 
maintained observation posts on the bar- 
rier islands but had not fortified them. 
The only mishap in this phase of Toe- 
nails occurred early in the morning of 
30 June, when the Zane ran on a reef 
while maneuvering in the badly charted 
waters in the rain. She was pulled free 
by the tug Rail in the afternoon. 

The landing of the i72d Infantry on 



Ships Moving Toward Rendova, late afternoon, 29 June 1943. 

Rendova was somewhat disorderly. C 
and G Companies, guided by Maj. Mar- 
tin Clemens and Lt. F. A. Rhoades, 
RAN, of the coastwatchers, and by native 
pilots, were to have landed from the de- 
stroyer-transports Dent and Waters on 
East and West Beaches of Rendova Har- 
bor at 0540 to cover the main body of 
the i72d Infantry when it came ashore. 22 
But again the weather played the Allies 
foul. The mist and rain obscured the 

22 C and G Companies, l^zd Infantry, and A Com- 
pany, 169th Infantry, had received special physical 
conditioning and training in jungle fighting and 
small boat handling. They were given the some- 
what romantic title of "Barracudas." Clemens, a 
former district officer in the government of the 
British Solomon Islands Protectorate, was a major 
in the British Solomon Islands Defense Force, and 
had been a great help during the Guadalcanal 
Campaign. Rhoades had been a plantation manager 
before the war. 

Renard Entrance markers and the white 
signal light on Bau Island that the recon- 
naissance party, present on Rendova 
since 16 June, had set up. As a result the 
APD's first landed C and G Companies 
several miles away, then had to re-em- 
bark them and go to the proper place. 

Meanwhile, the six transports took 
their stations north of Renard Entrance 
as the destroyers took screening posi- 
tions to the east and west. By now the 
clouds had begun to clear away, and vis- 
ibility improved. The troops gathered 
on the transport decks, and the first wave 
climbed into the landing craft at the 
rails, carrying their barracks bags with 
them. The order "All boats away, all 
troops away" was given aboard Turner's 
flagship, the transport McCawley, as the 
sun rose at 0642. Four minutes later 

Aboard the Transport McCawley, Admiral Turner's flagship, 29 June 1943- 
From left, Brig. Gen. Leonard F. Wing, Rear Adm. Theodore S. Wilkinson, Rear 
Adm. Richmond Kelly Turner, and Maj. Gen. John H. Hester. 

Turner warned the first boats as they 
headed for shore, some three thousand 
yards to the south: "You are the first 
to land, you are the first to land— expect 
opposition." 23 

As the landing craft moved shoreward 
the waves became disorganized. When 
the craft reached Renard Entrance be- 
tween Bau and Kokorana Islands, there 
was confusion and milling about until 
they began going through the entrance 
two abreast toward the narrow East 
and West Beaches that fronted Lever 
Brothers' 584-acre plantation. 

As the first landing craft touched down 
about 0700, the troops sprang out and 

CTF 31 War Diary, go Jun 43 entry. 

ran across the beaches into the cover of 
the jungle. C and G Companies reached 
Rendova Harbor about ten minutes after 
the troops from the transports, and they 
joined with the main body and moved 
inland toward the Japanese. 24 

25 43d Division documents give no data on the 
planned composition and timing o£ assault waves. 
They are at variance regarding the time of land- 
ing. The 43d Division and New Georgia Occupa- 
tion Force reports {which are virtually identical) 
state that C and G Companies landed at 0630 and 
that the main body began landing at 0745. These 
assertions are undoubtedly incorrect. The i72d re- 
port states that C and G Companies went astray 
but landed along with the main body about 0700. 
And the 43d Division's time o£ 0745 is contradicted 
by that division's G— 3 Journal which states that 
the division command post opened on Rendova at 
0730, CTF 31 War Diary states that the first troops 
hit the beach at 0656, 



The Japanese Rendova detachment— 
about 120 troops from the 229th Infantry 
and the Kure 6th Special Naval Landing 
Force— had been alerted early during the 
morning of go June. The alert proved 
to be a false alarm and they went back 
to sleep. The next alert— their first reali- 
zation that they were being attacked— 
came when the American assault craft 
hit the beach. As it was too late for the 
Japanese to man their beach defense 
positions, they posted themselves in the 
coconut plantation about one hundred 
yards behind East Beach. Radiomen tried 
to warn Munda but could not get the 
message through. A lookout at Banieta 
Point fired four blue flares and signaled 
headquarters by blinker. 

The Japanese could not hope to do 
more than harass the Americans. The 
special naval landing force commander, 
hit in the face by a burst from a BAR, 
was an early casualty. When about a 
dozen men were dead, the disorganized 
Japanese fell back into the jungle. They 
are reported to have lost some fifty or 
sixty men, while killing four Americans 
and wounding five, including Col. David 
M. N. Ross, the i72d's commander. By 
the end of the day the Americans had 
pushed inland one thousand yards. The 
105-mm. howitzers of the 103d Field 
Artillery Battalion were in position to 
cover Renard Entrance, the north coast 
of Rendova, and the barrier islands. 

All troops except working parties on 
board ship were ashore within thirty 
minutes after the landing of the first 
wave. This number included General 
Harmon who went along to observe 
operations. In the absence of strong en- 
emy resistance on Rendova, the chief 
problem that confronted the invaders 

was unloading supplies, getting them 
ashore, and moving them inland. Less 
than half an hour after they had been 
lowered into the water, the first landing 
craft returned to the ships for cargo. No 
landing waves were formed; each craft 
moved cargo ashore as soon as it was 
loaded by its mother ship. 

The first real delay in unloading was 
caused by shallow water. Many tank 
lighters (LCM's) grounded on reefs in 
the harbor and lost time refloating and 
finding passages through deeper water. 
Many lighters, grounding about fifty feet 
offshore, had to lower their ramps in 
water while the troops waded ashore with 
cargo in their hands or on their shoul- 
ders. In consequence, disorderly stacks 
of gear began piling up near the shore 
line. The beachmaster attempted, with 
only partial success, to prevent this. 

During most of the morning the Jap- 
anese did little. The Rendova garrison 
had not amounted to much; after the 
war the Japanese explained that the 
Munda and Rabaul commanders had not 
expected the Americans to land on the 
offshore islands. "Therefore," a post- 
war report states, "the landing on REN- 
DOVA Island completely baffled our 
forces." 25 When it became clear that the 
Americans were indeed landing on Ren- 
dova, 120-mm. and 140-mm. naval coast 
defense batteries at Munda and Baanga 
Island opened up on the ships, and they 
immediately replied with 5-inch fire. 
The destroyer Gwin was soon hit. She 
was the only casualty in the exchange of 
fire between ships and shore batteries 
that continued all day. But Turner and 
General Hester were operating very close 

" Southeast Area Naval Operations, II, Japanese 
Monogr No. 49 (OCMH), 36. 



Men of 43D Signal Company Wading Ashore from LCM's with signal 
equipment^ 30 June 194). 

to the Japanese air bases in southern 
Bougainville and the Shortlands, and 
these presented the greatest danger. For- 
tunately for the Americans, the Japanese 
were not prepared to counterattack at 

The commanders at New Georgia were 
not the only Japanese surprised by the 
invasion. Those at Rabaul were taken 
equally unaware. They had, of course, 
known that some form of Allied activity 
was impending in late June. The move 
to help Kennedy, the increasing tempo 
of Allied air and naval action, and inter- 
cepted Allied radio traffic told them as 
much. So Admiral Kusaka gathered air 
attack forces together and sent them to 
the airfields around Buin. But after 26 
June, when Allied movements seemed 

to slow down (Turner's task force was 
then rehearsing in the New Hebrides), 
the Japanese command concluded that 
the Allies had been simply reinforcing 
Guadalcanal on a grand scale. Kusaka 
pulled his air units back to Rabaul. Thus 
it was that although a submarine had 
sighted Turner's ships south of New 
Georgia about midnight, Kusaka, with 
sixty-six bombers, eighty-three fighters, 
and twenty reconnaissance seaplanes at 
his disposal, could do nothing about the 
invasion for several hours. 

Turner, whose plans called for unload- 
ing to be completed by 1 1 30, was first 
interrupted by a false air raid alarm at 
0856. The ships stopped unloading and 
steamed around in Blanche Channel 
while the thirty-two fighter planes cover- 



ing the landing got ready to intercept. 
The reported enemy planes failed to ap- 
pear, and unloading was resumed. 

The first real enemy air attack, a sweep 
by twenty-seven fighters, came just after 
1 1 oo. The Allied fighter cover shot most 
of them down before they could do any 
damage, but Turner's schedule was fur- 
ther delayed by the necessity for going 
to general quarters and getting under 

By about 1500 all but about fifty tons 
of gear had been unloaded. Turner or- 
dered the transports and screening de- 
stroyers back to Guadalcanal and they 
speedily took their departure. Shortly 
afterward twenty-five Japanese bombers, 
escorted by twenty-four fighters, came 
down from Rabaul. The majority of the 
bombers were shot down, but one man- 
aged to put a torpedo into the flagship 
McCawley. At 1715 eight more bombers 
struck at the retiring task force but 
failed to score. That evening overeager 
American PT boats, mistaking the crip- 
pled McCawley for an enemy, put two 
more torpedoes into her sides and she 
sank in Blanche Channel, fortunately 
without loss of life. 

Meanwhile the landing and handling 
of supplies on Rendova had been less 
than satisfactory. The invading forces 
had hoped to use Rendova Plantation 
to store supplies, although the preinva- 
sion patrols had not been able to inves- 
tigate it thoroughly because the Japa- 
nese were there. As the rain continued, 
the streams flooded, and the red clay of 
the plantation turned into mud. The 
mile-long prewar road that linked East 
and West Beaches served well early in 
the day, but soon heavy truck traffic 

ground it into a muddy mess. Seabee 
drivers of the 24th Naval Construction 
Battalion had to hook their truck cables 
to trees and winch their 21^-ton, 6x6 
trucks along in order to haul supplies 
from the heaped beaches to the cover 
and safety of high ground farther inland. 
They cut hundreds of coconut logs into 
twelve-foot lengths and tried to corduroy 
the roadbed, but the mud seemed to be 
bottomless. One bulldozer sank almost 
out of sight. To add to the supply diffi- 
culties, many containers were inade- 
quately marked and medical supplies be- 
came mixed among rations, fuel, and 
ammunition. The Rendova naval base 
force could not find all its radios, and 
little was known regarding the progress 
of operations at Wickham and Viru. The 
clutter and confusion caused by bogged 
trucks on the muddy roads and trails 
finally became so bad that the next day 
General Hester requested Turner to stop 
further shipments of trucks until the 
beachhead could be better organized. 

Despite the confusion ashore and the 
loss of the McCawley, operations on 30 
June were largely successful. Six thou- 
sand men of the 43d Division, the 24th 
Naval Construction Battalion and other 
naval units, and the gth Marine Defense 
Battalion had come ashore with weapons, 
rations, fuel, ammunition, construction 
equipment, and personal baggage. The 
Japanese had lost Rendova and several 
planes, and although they enthusiasti- 
cally reported inflicting heavy damage 
to Turner's ships, they admitted that, 
"due to tenacious interference by enemy 
fighter planes, a decisive blow could not 
be struck against the enemy landing 
convoy." "The speedy disembarkation of 



the enemy," they felt, "was absolutely 
miraculous." 20 

With the capture of the beachhead, 
General Hester dissolved the 17 2d Regi- 
mental Combat Team and returned the 
field artillery, engineers, and medical 
and communications men to divisional 
control. The build-up of troops and sup- 
plies for the attack against Munda and 
Vila was ready to begin. 

The second echelon of the Western 
Force came in on LST's the next day. 
This echelon included the 155-mm. how- 
itzers of the ig2d Field Artillery Bat- 
talion and the 155-mm. guns of A Bat- 
tery, gth Marine Defense Battalion. Suc- 
ceeding reinforcements continued to ar- 
rive at Rendova, Segi Point, Viru, and 
Wickham through 5 July until virtually 
the entire New Georgia Occupation 
Force as then constituted was present in 
New Georgia, with the main body at 

The Japanese were unable to do any- 
thing to prevent these movements, and 
did little damage to the beachhead. Only 
Japanese aircraft made anything like a 
sustained effort. Storms and poor visi- 
bility continued to prevent Allied planes 
from striking at the Shortlands-Bougain- 
ville fields, although they were able to 
hit the Munda and Vila airfields as well 
as Bairoko. The Japanese reinforced 
their air strength at Rabaul and sent 
planes forward to southern Bougainville 
and the Shortlands. On 2 July Admiral 
Kusaka had under his command 1 1 fight- 
ers and 13 dive bombers from the carrier 
Ryuho, 11 land-based twin-engine bomb- 
ers, 20 fighters, 2 reconnaissance planes, 
and a number of Army bombers that 

Ibid., pp. 29, 37. 

were temporarily assigned. The same day 
foul weather began closing in the rear- 
ward Allied bases. About noon the Com- 
mander, Aircraft, Solomons, from his 
post on Guadalcanal, ordered all Allied 
planes back. This left New Georgia with- 
out air cover. To make matters worse, 
the gth Marine Defense Battalion's SCR 
602 (a search radar designed for imme- 
diate use on beachheads) broke down 
that morning, and the SCR 270 (a long- 
range radar designed for relatively per- 
manent emplacement) was not yet set 

Kusaka sent all his planes to New 
Georgia. They reached the Rendova area 
in the afternoon, circled behind the 
clouded 3,448-foot twin peaks of Ren- 
dova Peak, then pounced to the attack. 
Many soldiers saw the planes but thought 
they were American until fragmentation 
clusters dropped by the bombers began 
exploding among them. The Rendova 
beachhead, with its dense concentra- 
tions of men and materiel, was an ex- 
cellent target. At least thirty men were 
killed and over two hundred were 
wounded. Many bombs struck the fuel 
dumps, the resulting fires caused fuel 
drums to explode, and these started more 
fires. Three 155-mm. guns of the gth 
Marine Defense Battalion were damaged. 
Much of the equipment of the 125-bed 
clearing station set up by the 1 18th Med- 
ical Battalion was destroyed; for a time 
only emergency medical treatment could 
be rendered. The wounded had to wait 
at least twenty-four hours before they 
could receive full treatment at Guadal- 

That night nine Japanese destroyers 
and one light cruiser shelled Rendova 
but hit nothing except jungle. The Jap- 



anese, it was clear, did not intend to land 
troops on Rendova, but they did not in- 
tend to allow the Americans to remain 
there unmolested. The air attacks, while 
serious, did not disrupt preparations for 
the next phase of Toenails. 

The Move to Zanana 

After the occupation of Rendova, the 
next tasks facing the invaders were the 
movement to the New Georgia mainland 
and the assault against Munda airfield. 
On 2 July Admiral Halsey, doubtless en- 
couraged by the lack of effective Japa- 
nese opposition, directed Turner to pro- 
ceed with plans for the move against 
Munda. To carry out these plans, Tur- 
ner on 28 June had reorganized the 
Western Force into five units: the trans- 
port unit consisting of destroyer-trans- 
ports and high-speed minesweepers; a 
destroyer screen; a fire support group, 
eventually consisting of three light cruis- 
ers and four destroyers; two tugs; and 
the Munda-Bairoko Occupation Force 
under General Hester. 

The Munda-Bairoko Occupation Force 
was further divided into five compo- 
nents. The Northern Landing Group, 
under Colonel Liversedge, was to oper- 
ate against Bairoko. The Southern Land- 
ing Group (the 43d Division less the 
1st Battalion of the 103d Infantry, the 
136th Field Artillery Battalion, the 9th 
Marine Defense Battalion less elements, 
and the South Pacific Scouts) under 
Brig. Gen. Leonard F. Wing, assistant 
commander of the 43d Division, was to 
attack Munda. The New Georgia Air 
Force, the Assault Flotillas (twelve LCI's, 
four LCT's, and native canoes), and a 
naval base group comprised the remain- 
ing three components. The Southern 

Landing Group was to land at Zanana 
Beach about five air-line miles east of 
Munda and attack westward to capture 
Munda while the Northern Group land- 
ed at Rice Anchorage in the Kula Gulf 
and advanced southward to capture or 
destroy the enemy in the Bairoko-Enogai 
area, block all trails from there to Mun- 
da, and cut off the Japanese route of 
reinforcement, supply, and escape. 27 

The troops on Rendova had been mak- 
ing ready since 30 June, but some of their 
efforts were marked by less than com- 
plete success. Hester had ordered ag- 
gressive reconnaissance of the entire area 
east and north of Munda. Starting on the 
night of 30 June-i July, patrols from 
the i^2d Infantry were to pass through 
Onaiavisi Entrance and Roviana Lagoon, 
land at Zanana, and begin reconnoiter- 
ing, while Marine patrols pushed south 
from Rice Anchorage. The 43d Division 
patrols were to operate from a base camp 
west of Zanana established on the after- 
noon of 30 June by Capt. E. C. D. Sher- 
rer, assistant intelligence officer of the 
New Georgia Occupation Force. 

At 2330, 30 June, despite a false rumor 
that Onaiavisi Entrance was impassable 
for small boats, patrols left Rendova on 
the eight-mile run to the mainland. The 
next morning regimental headquarters 
discovered that the patrols, unable to 
find the entrance in the dark, had landed 
on one of the barrier islands. The next 
evening the 1st Battalion, accompanied 
by Colonel Ross, shoved off for the main- 
land but could not find its way. Thus 
it was concluded that the move should 

21 Rice Anchorage lies about fifteen statute miles 
north by east of Munda Point. The other beach 
near Munda, Laiana Beach, lay within range of 
the Japanese artillery at Munda and would have 
been a risky place to land. 



Truck Towing a 155-MM. Howitzer Over Muddy Trail, Rendova, 7 July 1943. 

be made in daylight. Accordingly A 
Company, 169th Infantry, and the 1st 
Battalion, i72d Infantry, moved out for 
Zanana on the afternoon of 2 July. Na- 
tive guides in canoes marked the chan- 
nel. Everything went well except that 
about 150 men returned to Rendova at 
2330. Questioned about their startling 
reversal of course, they are reported to 
have stated that the coxswain of the lead- 
ing craft had received a note dropped 
by a B-24 which ordered them to turn 
back. 28 By the next morning, however, 
the entire 1st Battalion was on the main- 

The build-up of supplies on Rendova 
continued to be difficult; the rain and 

28 There seem to be no further available data re- 
garding this interesting but absurd excuse. 

mud partially thwarted the efforts of the 
118th Engineer and the 24th Naval Con- 
struction Battalions to drain the flat 
areas. East Beach was finally abandoned. 
The Occupation Force supply officers, 
after examining the solid coral subsur- 
faces under the sandy loam of the barrier 
islands, began using the islands as stag- 
ing points for supplies eventually intend- 
ed for the mainland. 

On the other hand, the artillery pic- 
ture was bright. General Barker, the 
artillery commander, had never planned 
to make extensive use of Rendova for 
artillery positions, as the range from 
Rendova to Munda was too great for 
all weapons except i55-mm. guns. Such 
barrier islands as Bau, Kokorana, Sasa- 
vele, and Baraulu could well support ar- 
tillery, and these islands, open on their 



north shores, possessed natural fields of 
fire. The field artillery could cover the 
entire area from Zanana to Munda, and 
initially would be firing at right angles 
to the axis of infantry advance and par- 
allel to the infantry front. This would 
enable the artillery to deliver extremely 
accurate supporting fire, since the dis- 
persion in artillery fire is greater in 
range than in deflection. On the other 
hand, it would increase the difficulty of 
co-ordination between artillery and in- 
fantry, for each artillery unit would re- 
quire exact information regarding not 
only the front line of the unit it was 
supporting, but also the front line of the 
unit's neighbors. Three battalions of ar- 
tillery were in place in time to cover the 
move of the ist Battalion, 173d Infan- 
try, to Zanana, and by 6 July two bat- 
talions of 105-mm. howitzers (the 103d 
and 169th), two battalions of 155-mm. 
howitzers (the 136th and ig2d), and two 
batteries of 155-mm. guns (gth Marine 
Defense Battalion) were in place, regis- 
tered, and ready to fire in support of the 

Antiaircraft managed to make a tre- 
mendous improvement over its perform- 
ance of 2 July, and celebrated Independ- 
ence Day in signal fashion when a close 
formation of sixteen unescorted enemy 
bombers flew over Rendova. This time 
radars were working, the warning had 
been given, and fire control men and 
gunners of the 9th Marine Defense Bat- 
talion's 90-mm, and 40-mm. batteries 
were ready. The Japanese flew into a 
concentration of fire from these weapons, 
and twelve immediately plunged earth- 
ward in return for the expenditure of 
eighty-eight rounds. The fighter cover 

from the Russells knocked down the re- 
maining four. 

Meanwhile, at Zanana, the 1st Bat- 
talion, i^2d Infantry, established a per- 
imeter of 400 yards' radius, wired in and 
protected by machine guns, 37-mm, anti- 
tank guns, and antiaircraft guns. Here 
General Wing set up the 43d Division 
command post, and to this perimeter 
came the remaining troops of the i^2d 
and 169th Infantry Regiments in eche- 
lons until 6 July when both regiments 
had been completely assembled. Ground 
reconnaissance by 43d Division soldiers, 
marines, and coastwatchers, aided after 
3 July by the 1st Company, South Pacific 
Scouts, under Capt. Charles W. H. Tripp 
of the New Zealand Army, was still 
being carried on. The advance westward 
was ready to begin. 

Rice Anchorage 

While 43d Division troops were es- 
tablishing themselves at Zanana, Colonel 
Liversedge's Northern Landing Group 
was boarding ships at Guadalcanal and 
making ready to cut the Japanese com- 
munications north of Munda. The 
Northern Landing Group was originally 
to have landed on 4 July, but the delays 
in getting a foothold at Zanana forced 
Turner to postpone the landing, and all 
other operations, for twenty-four hours. 

Because the Bairoko-Enogai area, the 
New Georgia terminus of the Japanese 
seaborne line of communications, was 
strongly held, and because preinvasion 
patrols had reported the Wharton River 
to be unfordable from the coast to a point 
about six thousand yards inland, Turner 
and Liversedge had decided to land at 
Rice Anchorage on the south bank of 
the river about six hundred yards in- 



land. 29 Supervised by Capt. Clay A. Boyd, 
USMC, and Flight Officer J. A. Corri- 
gan of the RAAF and the coastwatchers, 
native New Georgians cleared the land- 
ing beach and bivouac areas inland, and 
began hacking two trails from Rice An- 
chorage to Enogai to supplement the one 
track already in existence. 

The organization of Liversedge's 
Northern Landing Group was somewhat 
odd; the group consisted of three bat- 
talions from three different regiments. 
The 3d Battalions of the 145th and 148th 
Infantry .Regiments of the 37th Division 
and the 1st Raider Battalion, 1st Marine 
Raider Regiment, made up the force. 30 
And the force was lightly equipped. In 
order to permit rapid movement through 
the thick jungles and swamps of the area 
north of Munda, the troops took no ar- 
tillery of any kind. Machine guns and 
mortars were their heaviest organic sup- 
porting weapons. 

The battalions boarded the APD's, 
destroyers, and minesweepers at Guadal- 
canal on the afternoon of 4 July. The 
troops carried one unit of fire and ra- 
tions for three days; five days' rations 
and one unit of fire were stowed as cargo. 
Escorted by Rear Adm. Walden L. Ains- 
worth's three light cruisers and nine de- 
stroyers, the speedy convoy started up 
the Slot at dusk. Shortly before midnight 
of a dark, rainy night, the ships rounded 
Visuvisu Point and entered Kula Gulf. 
Ains worth bombarded Vila and then 

"The patrols had left Segi on 14 June by boat. 
Turner had also considered landing the force at 
Roviana Lagoon and having it march overland to 
Bairoko but decided against it because the terrain 
was too rugged. 

30 Turner had originally planned to use the 4th 
Raider Battalion but when it was delayed at Vim 
Harbor the 3d Battalion, 145th, was substituted. 

Bairoko Harbour with 6-inch and 5-inch 
shells, while the transport group headed 
for Rice Anchorage. 

As the cruisers and destroyers were 
concluding their bombardment, the de- 
stroyer Ralph Talbot's radar picked up 
two surface targets as they were leaving 
the gulf. These were two of three Jap- 
anese destroyers which had brought the 
first echelon of four thousand Japanese 
Army reinforcements down from the 
Shortland Islands. 31 The Japanese ships 
had entered Kula at the same time as 
Ainsworth; warned by his bombardment, 
they were clearing out, but fired tor- 
pedoes at long range. One scored a fatal 
hit on the destroyer Strong. As two other 
destroyers were taking off her crew, four 
140-mm. Japanese seacoast guns at Eno- 
gai opened fire, joined soon by the Bai- 
roko batteries, but did no damage. 32 

Liversedge's landing started about 
0130, just after Ainsworth's bombard- 
ment ceased. The APD's unloaded first, 
then destroyers, finally minesweepers. 
Each LCP (R) towed one ten-man rub- 
ber boat to shore. The way was marked 
by native canoes and shore beacons. The 
Japanese batteries harassed the troops but 
did not hit anything. There were no 
Japanese on the landing beach. 

Nonetheless the landing was attended 
by troubles. A shallow bar obstructed 
the mouth of the Wharton River so ef- 
fectively that many boats were grounded 
and later craft got over the bar only by 
coming in with lighter loads. The land- 
ing beach was too small to accommo- 

31 See below. 

32 Ainsworth 

had first wanted to bombard Enogai 
but did not because air reconnaissance showed no 
evidence of shore batteries there. 



date more than four boats at once, and 
the river mouth was thus continually 
jammed with loaded boats waiting their 
turns at the beach. Also, about two hun- 
dred men of the 3d Battalion, 148th In- 
fantry, were landed at Kobukobu Inlet, 
several hundred yards north of Rice 
Anchorage, a mishap which may have 
occurred because of the darkness of the 
night. Some days elapsed before the two 
hundred men made their way through 
the jungle to catch up with their bat- 

As dawn of 5 July was breaking, the 
volume of fire from the Enogai batteries 
against the ships was increasing, and it 
seemed unwise to risk this fire in day- 
light as well as to invite air attack. All 
but seventy-two troops and 2 percent of 
the cargo had been put ashore. There- 
fore, the convoy commander withdrew. 
Liversedge, with nearly all his three bat- 
talions ashore and under his control, 
made ready to move south. 

Thus by 5 July Toenails was over. 
Throughout the complicated series of 
operations certain characteristics stood 
out. The weather had been consistently 
foul. The Japanese had not been able 
to resist effectively. The American per- 
formance, in spite of several instances of 
confusion, was very good, in that six 
landings in all had been carried out ac- 
cording to a complicated schedule that 
called for the most careful co-ordination 
of all forces. Clearly, Admiral Turner's 
reputation as an amphibious commander 
was well founded. 

The Americans had now established 
themselves in New Georgia. Viru Har- 
bor and Wickham Anchorage were se- 
cure points on the line of communica- 
tions. The airfield at Segi Point was 
nearing completion. And at Rice An- 
chorage and Zanana General Hester's 
Munda-Bairoko Occupation Force was 
making ready to strike against Munda 


The Offensive Stalls 

Although enemy resistance had been 
ineffective, and casualties in Toenails 
were relatively few, the Japanese were 
not finished. They planned to hold New 
Georgia. The New Georgia Occupation 
Force had had difficulties, but greater 
troubles were in store for it. 

Japanese Plans 

On 2 July, with the Americans in pos- 
session of Rendova, Segi Point, and Viru 
Harbor, the Japanese altered their com- 
mand on New Georgia. 1 By mutual 
agreement Maj. Gen. Noboru Sasaki, 
commander of the Southeastern Detach- 

1 Unless otherwise indicated this chapter is based 
on SOPACBACOM, History of the New Georgia 
Campaign, Vol. I, Ch. Ill, OCMH; the jnls, diaries, 
and after action rpts o£ COMSOPAC, CTF 31, 
NGOF, XIV Corps, 43d Div, 43d Div Arty, 1st Mar 
Raider Regt, 145th Inf, 148th Inf, 169th Inf, and 
i72d Inf; 8th Area Army Operations, Japanese 
Monogr No. 110 (OCMH); 17th Army Operations, 
Vol. II, Japanese Monogr No. 40 (OCMH); South- 
east Area Naval Operations, Vol. II, Japanese 
Monogr No. 49 (OCMH); Outline of Southeast 
Area Naval Air Operations, Pt. IV, Japanese 
Monogr No. 108 (OCMH); Operations of the 1st 
Battalion, 169th Infantry (43d Infantry Division) 
in the New Georgia Campaign: 30 June— 18 July 
1943 (Northern Solomons Campaign), a monograph 
relating the personal experience of a battalion in- 
telligence officer, prepared by Maj. Jack Swaim; Ltr, 
Lt Col Marvin D. Girardeau to Chief of Military 
History, sub; Comments Re Hist Monogr, Marines 
in Central Solomons, 6 Feb 57, with inclosures, 

merit, took over direction of all Army 
and Navy forces in New Georgia, This 
action brought Rear Adm. Minoru Ota's 
8th Combined Special Naval Landing 
Force under Sasaki, who was under the 
tactical control of the 8th Fleet. Except 
for small detachments on Vella Lavella, 
Gizo, and other islands, the 10,500 men 
in Sasaki's joint force were about evenly 
divided between Kolombangara and 
Munda. At Kolombangara, under Col. 
Satoshi Tomonari, were two battalions 
of the 13th Infantry, most of the 3d Bat- 
talion, 22pth Infantry, the Yokosuka jth 
Special Naval Landing Force (less ele- 
ments), and artillery and engineer units. 
Guarding Munda, where Sasaki and Ota 
maintained their headquarters, were Col. 
Genjiro Hirata's izyth Infantry (less 
two battalions) and artillery, engineer, 
communication, and medical units. The 
main body of the Kure 6th Special Naval 
Landing Force was concentrated at Bai- 

Sasaki was well aware that the Amer- 
icans would attack Munda. He could 
see the troops moving from Rendova to 
the mainland. Munda field was receiving 
shellfire from the American 155's. If fur- 
ther proof was needed, Japanese patrols 
had brushed with the Allies near Zanana 
on 3 July, and the next day the 229th 
Infantry reported a clash with about five 



Maj. Gen. Noboru Sasaki 

hundred Americans in the same place. 
Immediately after the invasion of Ren- 
dova Sasaki had instructed Tomonari to 
alert his units for possible transfer to 
Munda and directed that two 140-mm. 
naval guns and two mountain guns be 
moved from the Bairoko area to Munda. 
After receiving the 22gth's report he 
brought the 3d Battalion, 229th Infan- 
try, from Kolombangara through Bai- 
roko to Munda to rejoin the regiment 
on 4 July. 

On the same day, Sasaki proposed a 
counterlanding against Rendova. As their 
artillery pieces lacked the range to hit 
Rendova, the Japanese on Munda could 
not retaliate when shells from American 
155's crashed on Munda field. Sasaki 
therefore suggested that the main body 
of the Munda garrison board landing 

craft, avoid recognition by mingling with 
American craft, and assault Rendova 
amid the resulting confusion. This in- 
teresting plan might have succeeded and 
caused a disaster to the Allies. More 
probably, by removing the Munda troops 
from their strong defense positions, it 
would have saved the Americans a lot 
of fighting. 8th Fleet Headquarters ap- 
parently vetoed the proposal. 

Also on Independence Day General 
Imamura and Admiral Kusaka, who 
wished to hold New Georgia at all costs 
as a key outpost for Bougainville, con- 
sidered the problem of holding the 
island in relation to the general defense 
of the Southeast Area. They decided to 
strengthen New Georgia and to hold 
New Guinea with the troops already 
there. Imamura agreed to give four thou- 
sand more ijth Army troops to Sasaki. 
These, including additional units from 
the 13th and 229th Infantry Regiments 
plus artillerymen, engineers, and med- 
ical men, were to be shipped in echelons 
from Erventa in the Shortlands to Ko- 
lombangara. Warships would transport 
them. It was the first echelon of these 
troops that Admiral Ainsworth's task 
force kept from landing on the night of 

4-5 J u 'y- 

On 5 July the Japanese naval offi- 
cers' worries regarding New Georgia 
were increased by Hester's build-up at 
Zanana and Liversedge's landing at Rice 
Anchorage. The Japanese assigned ten 
destroyers to transport the second eche- 
lon, which was to be put ashore at Vila 
in the early morning hours of 6 July. 
Informed that Japanese warships were 
getting ready to sail from the Shortlands, 
Halsey ordered Ainsworth's task group 
to intercept, reinforced by two destroy- 



ers to replace the Strong and the dam- 
aged destroyer Chevalier. Ainsworth, re- 
tiring from the Kula Gulf, was in In- 
dispensable Strait when Halsey's orders 
reached him. He reversed course and 
entered Kula Gulf about midnight, a 
few minutes behind the Japanese de- 
stroyers. In the ensuing Battle of Kula 
Gulf, the veteran cruiser Helena was 
sunk. The Japanese lost the destroyers 
Niizuki and Nagatsuki, but put 850 
soldiers ashore at Vila. 2 This addition of 
850 men enabled Sasaki to send part of 
another battalion from Kolombangara to 
Munda that same day. 

Admiral Kusaka, who moved his head- 
quarters from Rabaul to Buin "to alter 
the grave situation and raise the morale 
of all the forces," wanted still more 
troops for New Georgia. 3 On 7 July he 
asked Imamura for 11,000 more soldiers. 
The general, who had just approved 
sending 4,000 men to New Georgia, now 
stated that he doubted that even Bou- 
gainville could be made secure. Al- 
though willing to consider sending an- 
other division to Bougainville, he re- 
fused to provide 11,000 more troops for 
New Georgia. 

It was well for the Americans that 
Imamura refused the 11,000 men. Blast- 
ing the existing garrisons out of Munda 
and Bairoko was to prove sufficiently 
difficult and bloody. 

Operations of the Northern Landing 

The March to Dragons Peninsula 
At 0600 of 5 July, with nearly all his 

Northern Landing Group ashore and in 
hand, Colonel Liversedge ordered his 
troops to move out. The 1st Marine 
Raider Battalion, the 3d Battalion, 148th 
Infantry, and K and L Companies of the 
145th Infantry were to advance south- 
ward toward Dragons Peninsula, the 
piece of land lying betw een Enoga i Inlet 

! For a full account see Morison, Breaking the 
Bismarcks Barrier, pp. 160-73. 

3 Southeast Area Naval Operations, II, Japanese 
Monogr No. 49 (OCMH), 32. 

and Bairoko Harbour. (Map 8) Once 
they had reached the head of Enogai 
Inlet, the Raiders and K and I Com- 
panies, 145th, were to swing right to 
take the west shore of Enogai Inlet prior 
to assaulting Bairoko, while the 3d Bat- 
talion, 148th, advanced southwest to 
block the Munda-Bairoko trail. M, L, 
and Headquarters Companies of the 3d 
Battalion, 145th Infantry, were ordered 
to stay and defend Rice Anchorage un- 
der Lt. Col. George C. Freer, the bat- 
talion commander. 

The preinvasion reconnaissance par- 
ties, after examining the ground between 
Rice Anchorage and Dragons Peninsula 
to determine whether an overland at- 
tack would be practicable, had reported 
the country generally level with sparse 
undergrowth. There were no swamps. 
Enogai Inlet, with a good anchorage, had 
a mangrove-covered shore line except at 
its head where firm ground rose steeply 
to an elevation of about five hundred 
feet. Dragons Peninsula itself was hilly, 
swampy, and jungled, but on the inland 
shore of Leland Lagoon a ridge ran from 
Enogai to Bairoko village. Bairoko Har- 
bour was deep, and was backed by 

The advance to Dragons Peninsula be- 
gan immediately after Liversedge issued 
his orders. Guided by natives, the troops 
moved along the three parallel trails— 
the original track and the two cut by 



Corrigan's natives. The ist Marine Raid- 
er Battalion, Lt. Col. Samuel B. Grif- 
fith, III, commanding, led the way, fol- 
lowed in order by K and L Companies, 
145th, under Maj. Marvin D. Girardeau, 
and Lt. Col. Delbert E. Schultz's 3d 
Battalion, 148th Infantry. The patrols' 
reports had implied that the going would 
be easy, but the march proved difficult. 
The rough trails, winding over hills and 
ridges, were obstructed by branches, 
roots, and coral outcroppings. Rain wet 
the troops all day. The Raiders, whose 
heaviest weapon was the 60-mm. mor- 
tar, made fairly steady progress, but the 
soldiers of M Company, 148th Infantry, 
fell behind as they floundered through 
the mud with their heavy machine guns, 
81 -mm. mortars, and ammunition. 

At 1300 part of D Company of the 
Raiders, the advance guard, was sent on 
ahead to secure a bridgehead on the far 
bank of the Giza Giza river. 4 Three 
hours later the Raiders' main body and 
the companies of the 145th Infantry ar- 
rived at the river and bivouacked there 
overnight. They had covered five and 
one-half miles in the day's march with- 
out meeting a single Japanese. Colonel 
Schultz's battalion camped for the night 
about one and one-half miles to the 

Next morning, 6 July, the Raiders led 
out again, and D Company pushed ahead 
to secure a crossing over the Tamakau 
River. The rains had flooded the river; 
it was now nine feet deep. Without tools 
or time to build a proper bridge, the 
Raiders threw a log over the stream, and 
improvised rafts from poles and ponchos 

4 At this time the lettered companies of the ist 
Raider Battalion were all rifle companies; there was 
no heavy weapons company. 

to ferry over their heavy equipment. 
After several rafts had capsized, they gave 
up and carried everything over on the 
log. Several men slipped off the log and 
fell into the swollen river; a few had to 
be rescued from drowning. The crossing 
had started before noon, but not until 
dusk did the last man cross the river. 
Schultz's battalion, also delayed by high 
water, caught up with the Raiders and 
bivouacked near them for the night. 

On the morning of 7 July the Raiders 
and Girardeau's companies set out for 
Enogai, while Schultz's battalion pushed 
south toward the Munda-Bairoko trail. 
The country was rough, the going hard 
for both forces. The Raiders took five 
hours to cover the 2,500 yards between 
their bivouac and the east end of Enogai 

The 3d Battalion, 148th, reached the 
trail at 1700. In the afternoon the two 
hundred men who had been landed 
astray on 5 July caught up with the main 
body. There had been no opposition 
from the Japanese; a patrol was ob- 
served but kept its distance. 

Capture of Enogai Inlet 

When the Marine Raiders and Gi- 
rardeau's two companies reached Enogai 
Inlet, one platoon, again from D Com- 
pany, pushed forward to secure the de- 
serted village called Maranusa. From 
there a patrol marched toward Triri, 
another village which was hardly more 
than a clearing. Up to now the marines 
had not seen any Japanese, but as the 
patrol approached Triri its point de- 
tected five Japanese ahead. The marines 
ambushed the party and killed two of its 
members. They belonged to the Kure 
6th Special Naval Landing Force. The 



other three fled. When Liversedge heard 
about this action, which made it obvious 
that his force had been discovered, he 
ordered Griffith to secure Triri at once 
in order to prepare to repel a counter- 
attack. Griffith dispatched the demoli- 
tion platoon from battalion Headquar- 
ters Company with orders to pass through 
D Company and seize Triri. On the way 
up, the platoon ran into a strong enemy 
patrol which opened fire. The marines 
retired to a defensive position on the 
bank of a stream and kept the enemy 
in place with fire. At this point D Com- 
pany appeared on the scene, swung to 
the left, struck the Japanese on their 
inland (right) flank, and drove them off. 
Three marines and ten Japanese were 
killed in this skirmish. One of the dead 
Japanese had on his person a defense 
plan which showed the exact location of 
the heavy guns at Enogai. By 1600 all 
elements of the Enogai attacking force 
were installed at Triri. 

At dawn the following morning— 8 
July, the day on which Schultz's bat- 
talion completed its block on the 
Munda-Bairoko trail— two Raider com- 
bat patrols went out of Triri. B Com- 
pany sent one out to ambush a trail 
which led northwesterly to Enogai, and 
a D Company patrol advanced south 
along a cross-country track leading to 
Bairoko to lay another ambush. This pa- 
trol had advanced a short distance by 
0700, when it ran into an enemy force 
of about company strength. A fire fight 
broke out, and at 1000 Griffith sent C 
Company to drive the enemy back a 
short distance. 

In the meantime, the patrol which had 
advanced toward Enogai reported no 
contact with the enemy. In order to as- 

semble all companies of the 1st Raider 
Battalion for the attack against Enogai, 
Griffith sent K and L Companies of the 
145th south to take over from C Com- 
pany. C Company then disengaged, 
moved back to Triri, and in the early 
afternoon the 1st Raider Battalion 
marched northwest toward Enogai. But 
the trail led the marines into an im- 
passable mangrove swamp. The bat- 
talion therefore retired to Triri, while 
scouts hunted for a better route to use 
the next day. 

In the south sector, the fight between 
the Japanese and K and L Companies 
had continued. The Japanese in repeat- 
ed assaults struck hard at K Company 
which was on the right (west). Capt. 
Donald W. Fouse, commanding K Com- 
pany, was wounded early in the action 
but stayed with his company until the 
fight was over. When the Raider bat- 
talion retired to Triri, the Demolition 
Platoon was committed to the line, and 
when K Company was hard hit a platoon 
from B Company of the Raiders swung 
wide around the Japanese left flank and 
struck them in the rear. This maneuver 
succeeded. The enemy scattered. 5 

The 1st Raider Battalion resumed its 
advance against Enogai the next morn- 
ing, using a good trail, apparently un- 
known to the Japanese, that led over 
high ground west of the swamp. K and 
L Companies remained to hold Triri, 
the site of Liversedge's command post. 
Griffith's battalion, meeting no opposi- 
tion, made good time. By 1100 the ma- 
rines were in sight of Leland Lagoon. 
They swung slightly to the right toward 

B K Company reported killing a hundred Japa- 
nese; the marine platoon is reported to have killed 



Enogai Point and at 1500 ran into two 
Japanese light machine guns which 
opened fire and halted the advance. Grif- 
fith deployed, with A Company on the 
left, C in the center, B on the right, and 
D in reserve. The companies then as- 
saulted, but the Japanese defended so 
resolutely that no further progress was 
made that day. 

Patrols reconnoitered vigorously so 
that by 0700, 10 July, Griffith had been 
informed that the Japanese were strong- 
est in front of his center and left, and 
that there were no Japanese directly in 
front of B Company. The battalion re- 
sumed the attack at 0700. C and A 
Companies advanced slowly against rifle 
and machine gun fire. Supported by 60- 
mm. mortars, B Company drove forward 
rapidly, cleared the village of Baekineru, 
and captured two machine guns. Then 
A Company, strengthened by one pla- 
toon from battalion reserve, pushed over 
Enogai Point to the sea. By 1500 all 
organized resistance had ended except 
for a pocket in front of A Company. 
When D Company started establishing 
beach defenses, it was troubled by three 
machine guns from another enemy 
pocket. Mopping up these two groups 
of Japanese took until 1 1 July. 

The Raiders had run out of food and 
water by midafternoon of 10 July, but 
were succored by L Company, 145th, 
which brought rations and water up 
from Triri. These had been dropped, at 
Liversedge's urgent request, by C-47's 
from Guadalcanal. 

By 12 July Enogai was organized for 
defense against land or seaborne attacks. 
Estimates of Japanese casualties ranged 
from 150 to 350. Postwar Japanese ac- 
counts assert that Enogai was defended 

by one platoon of soldiers and 81 men 
of the Kure 6th Special Naval Landing 
Force. The marines lost 47 killed, 4 miss- 
ing, and 74 wounded. They captured 3 
.50-caliber antiaircraft machine guns, 4 
heavy and 14 light machine guns, a 
searchlight, rifles, mortars, ammunition, 
2 tractors, some stores and documents, 
and the 4 140-mm. coastal guns that had 
harassed the landing at Rice Anchorage. 
The guns were intact except that their 
breechblocks had been removed. Luck- 
ily, a marine digging a foxhole uncov- 
ered one, and a hasty search of the area 
turned up the other three. The marines 
used these guns to help guard the sea- 
ward approaches to the newly won posi- 
tion. 6 

Roadblock North of Munda 

While the Raiders were thus engaged, 
the soldiers of the 3d Battalion, 148th 
Infantry, were deep in the jungle holding 
their block. The block, completed on 8 
July, was set up on a well-used trail 
some two miles southeast of Enogai Inlet 
and eight miles north of Munda. I Com- 
pany, with one M Company platoon at- 
tached, faced toward Bairoko; K Com- 
pany faced Munda. L Company covered 
the flanks of I and K, and extended its 
lines back to protect the battalion com- 
mand post. M Company, with the Anti- 
tank Platoon attached, was in a sup- 
porting position to the rear. Each rifle 
company held one platoon in reserve 
under battalion control. All positions 
were camouflaged. Colonel Schultz or- 
dered his men to fire at enemy groups 
larger than four men; smaller parties 
were to be killed with bayonets. 

6 Sasaki apparently had ordered two of these guns 
to Munda. 



The battalion held the block from 8 
through 17 July. Patrols went out regu- 
larly. General Hester had ordered patrols 
to push far enough to the south to make 
contact with the 43d Division's right 
flank as it advanced westward against 
Munda, but this was never done. 

Schultz was strengthened on 1 1 July 
by the addition of I Company, 145th In- 
fantry, after a group of Japanese had 
overrun part of L Company's positions 
in a series of attacks starting 10 July. 
Except for this, the Japanese made no 
effort to dislodge Schultz's men, whose 
greatest enemy proved to be hunger. The 
troops had left Rice Anchorage carrying 
rations for three days on the assumption 
that Enogai Inlet would be taken in two 
days and that American vessels could 
then land stores there. These could be 
delivered to the troops after a relatively 
short overland haul. But Enogai was not 
secured until 11 July. The 120 native 
carriers thus had to carry food all the 
way from Rice Anchorage. Although, ac- 
cording to Colonel Liversedge, the na- 
tives "accomplished an almost superhu- 
man task," they could not carry supplies 
fast enough to keep the troops fed. 7 

By 9 July the food shortage was seri- 
ous. Only 2,200 D rations had been 
delivered. Late that evening, with food 
for the next day reduced to one ninth 
of a D and one ninth of a K ration per 
man, Schultz radioed to Liversedge an 
urgent request that food be brought in 
by carrier. He also hoped the natives 
could carry out two badly wounded men 
who were being cared for in the battalion 
aid station. But as there were not enough 
natives, C-47's dropped food, as well as 

' 1st Mar Raider Regt [NLG], Combat Rpt and 
War Diary, 4 Jul-29 Aug 43 entries. 

ammunition, to the battalion the next 
afternoon. Much of the food fell far be- 
yond the 3d Battalion's lines, and some 
of the ammunition was defective. Schultz 
was forced to cut the food allowance for 
the next twenty-four hours to one twelfth 
of a K ration. Fortunately Enogai had 
now fallen, and on 13 July Flight Officer 
Corrigan's natives carried in three hun- 
dred pounds of rice which the men 
cooked in their helmets, using salt tab- 
lets for seasoning. The next day, though, 
was another hungry one; one D and one 
K ration was the allowance for each 
eighteen men. Thereafter, until the block 
was abandoned, carrying parties and air 
drops kept food stocks high enough. 

During the nine days it held the block, 
the 3d Battalion lost 11 men killed and 
29 wounded; it estimated it had killed 
250 Japanese. 

At the time it was believed that the 
blockers had cut off Munda from rein- 
forcement via Bairoko, and that they 
held the Japanese Bairoko force in place, 
prevented Enogai from being reinforced 
from either Munda or Bairoko, and pro- 
tected Griffith's right flank and rear. 8 
Knowledge gained after the event indi- 
cates that none of these beliefs was 

That Munda was not isolated is dem- 
onstrated by the fact that the Japanese 
reinforcement of Munda was in full 
swing, and all the reinforcements seem 
to have reached Munda without much 
trouble. The enemy obviously stopped 
using the blocked trail after 8 July and 
shifted to another one farther west. 

Meanwhile, reinforcement by water 

8 NGOF, Report of Operations on New Georgia; 
1st Mar Raider Regt, Special Action Rpt, New 
Georgia, 6 Oct 43. 



continued. On 9 July, when 1,200 Japa- 
nese from the Shortlands landed on Ko- 
lombangara, 1,300 of the 13th Infantry 
transferred by barge to Bairoko. Three 
days later, on 12 July, a Japanese ten- 
ship force left Rabaul to carry 1,200 
more soldiers to Kolombangara, and Hal- 
sey sent Ainsworth's task force to inter- 
cept again. The two forces collided early 
on 13 July northeast of Kolombangara 
in a battle named for that island. The 
Allies lost the destroyer Gwin; the New 
Zealand light cruiser Leander and the 
American light cruisers St. Louis and 
Honolulu suffered damage. The Japa- 
nese flagship, the light cruiser Jintsu, was 
sunk, but 1,200 enemy soldiers were 
landed on the west coast of Kolomban- 
gara. 9 

At Bairoko, during this period, the 
13th Infantry made ready to go to Mun- 
da. It was part of this regiment which 
attacked the trail block on 10 July. 10 On 
13 July, when the Bairoko garrison was 
strengthened by the 2d Battalion, 45th 
Infantry, and a battery of artillery, the 
13th Infantry marched south to the 
Munda front. 

As far as pinning down the Bairoko 
troops was concerned, the block lay more 
than two miles from Bairoko, and thus 
could not have affected the Bairoko gar- 
rison very much. And surely, had the 
Japanese desired to reinforce Enogai 
from Bairoko, they would have used the 
direct trail along the shore of Leland 
Lagoon rather than going over the more 
roundabout route which was blocked. 

* For a full account of the battle see Morison, 
Breaking the Bismarcks Barrier, pp. 180—91. 

10 An enemy account claims that the Americans 
were "annihilated." See 17th Army Operations, Vol. 
II, Japanese Monogr No. 40 (OCMH). 

In view of the American strength at 
Rendova and Zanana, the thesis that the 
Japanese might have sent troops from 
Munda to Enogai is equally untenable, 
even if it were not known that the Jap- 
anese were reinforcing Munda, not Eno- 
gai. Finally, Schultz's battalion was too 
far from Griffith's to render much flank 
protection in that dense, dark jungle. 11 
It is clear that the trail block failed to 
achieve results proportionate to the ef- 
fort expended. So far, the principal ef- 
fect of the entire Rice Anchorage-Eno- 
gai-Bairoko operation had been to em- 
ploy troops that could have been better 
used at Munda. 

By 1 1 July, with Enogai secured, Liv- 
ersedge was five days behind schedule. 
Casualties, illness, and physical exhaus- 
tion had reduced the ist Raider Bat- 
talion to one-half its effective strength. 
Considering that two fresh battalions 
could reduce Bairoko in three days, he 
asked Admiral Turner, with Hester's ap- 
proval, for additional troops. There were 
not two more battalions to be had, Tur- 
ner replied, but he promised to land the 
4th Raider Battalion at Enogai by 18 
July, and authorized a delay in the as- 
sault against Bairoko until then. Thus 
short one battalion, Liversedge directed 
Schultz to abandon his block and march 
to Triri on 17 July. The 3d Battalion, 
148th, was to join the Raiders and part 
of the 3d Battalion, 145th, in the Bai- 
roko attack. 

The Northern Landing Group had ac- 
complished the first phase of its mission 
by capturing Enogai, but was behind 
schedule. On the Munda front, General 

11 On 10 July General Hester explicitly directed 
Colonel Liversedge to keep his battalions within 
supporting distance of one another. 



Wing's Southern Landing Group was also 
behind schedule. 

Operations of the Southern Landing 

From Zanana to the Barike River 

Once the 169th and i72d Regiments 
had landed at Zanana, General Hester 
had originally planned, the two regi- 
ments were to march overland about two 
and one-half to three miles to a line of 
departure lying generally along the Ba- 
rike River, then deploy and attack west 

to capture Munda airfield. (Map 9) The 
regiments were directed to reach the line 
of departure and attack by 7 July, but 
by the time the two regiments had 
reached Zanana all operations were post- 
poned one day. 

The overland approach to Munda in- 
volved a march through the rough, jun- 
gled, swampy ground typical of New 
Georgia. The terrain between Zanana 
and Munda was rugged, tangled, and pat- 
ternless. Rocky hills thrust upward from 
two to three hundred feet above sea level, 
with valleys, draws, and stream beds in 
between. The hills and ridges sprawled 
and bent in all directions. The map 
used for the operation was a photomap 
based on air photography. It snowed the 
coast line and Munda airfield clearly, 
but did not give any accurate indication 
of ground contour. About all the troops 
could tell by looking at it was that the 
ground was covered by jungle. 

The difficulty of travel in this rough 
country was greatly increased by heat, 
mud, undergrowth, and hills. Visibility 
was limited to a few yards. There were 
no roads, but a short distance north of 
Zanana lay Munda Trail, a narrow foot 
track that hit the coast at Lambeti Plan- 

tation. Engineers were making ready to 
build a road from Zanana to Munda 
Trail, and to improve the latter so that 
it could carry motor traffic. 

Having made their way from Zanana 
to the line of departure on the Barike, 
the two regiments would, according to 
Hester's orders, deliver a co-ordinated 
attack against Munda airfield, which lay 
about two and one-half miles westward. 
The i72d Infantry on the left (south) 
would be responsible for a front extend- 
ing inland from the coast. The 169th 
Infantry's zone of action lay north of the 
17 ad's; its right flank would be in the 
air except for protection given it by 
South Pacific Scout patrols operating to 
the north. The attack would be support- 
ed by General Barker's artillery and by 
air and naval bombardments. 

Two days after the beginning of the 
two-regiment attack, a heavy naval bom- 
bardment would prepare the way for an 
assault landing by the 3d Battalion, 103d 
Infantry, and the gth Marine Defense 
Battalion's Tank Platoon at 0420, 9 
July, at the west tip of Munda Point. 

Hester and Wing did not expect to 
meet any serious opposition between 
Zanana and the Barike River, and their 
expectations must have been confirmed 
by the experience of the 1st Battalion, 
i72d. On 3 July Colonel Ross had or- 
dered this battalion to remain at Zanana, 
making every effort at concealment. The 
message was apparently not received, for 
on 4 July the battalion, accompanied by 
A Company, 169th Infantry, easily 
marched to the Barike River, meeting 
only small Japanese patrols on the way. 
It was this premature move that helped 
to alert Sasaki. 

Next morning Captain Sherrer of the 

F. Tmmple 

MAP 9 



G-2 Section led a patrol of six New Zea- 
landers, twelve Americans, and eighteen 
Fijians from his base camp toward the 
upper reaches of the Barike River. They 
intended to set up a patrol base on high 
ground suitable for good radio trans- 
mission and reception. Normally they 
would have avoided detection by mov- 
ing off the trails and striking out through 
the wilderness, but, laden with radio 
gear, they followed Munda Trail. As 
they approached a small rise that lay 
about two miles from Zanana, and about 
eleven hundred yards east of the line of 
departure, they met enemy machine gun 
fire. They replied with small arms, and 
the fire fight lasted until dusk when 
Sherrer's group disengaged and went 
south to the bivouac of the ist Battalion, 
i72d, near the mouth of the Barike. B 
Company, 17 2d, went out to investigate 
the situation the next morning and 
found the Japanese still occupying the 
high ground, astride the trail. Attacks 
by B Company and by A Company, 
169th, failed to dislodge the Japanese. 
By afternoon of 6 July, however, the 
three battalions of the i72d Infantry 
were safely in place on the Barike, the 
1st and 3d on the left and right, the 2d 
in regimental reserve. 

But the 169th Infantry, commanded 
by Col. John D. Eason, was not so for- 
tunate. That regiment's 3d Battalion, 
under Lt. Col. William A. Stebbins, set 
out along the trail from Zanana to the 
line of departure on the morning of 6 
July. Natives guided the battalion as it 
moved in column of companies, each 
company in column of platoons, along 
the narrow trail. The men hacked vines 
and undergrowth to make their way 
more easily. Shortly after noon, General 

Wing ordered Stebbins' battalion to de- 
stroy the point of Japanese resistance 
that Sherrer had run into. 

It was estimated, correctly, that about 
one platoon was trying to block the trail . 
General Sasaki, aware of the Allied ac- 
tivity east of him, had ordered part of 
the nth Company, 229th Infantry, to 
reconnoiter the Barike area, clear fire 
lanes, and establish this trail block with 
felled trees and barbed wire. 

The 3d Battalion, 169th, apparently 
did not run into the block on 6 July. It 
dug in for the night somewhere east of 
the block, but does not seem to have es- 
tablished the sort of perimeter defense 
that was necessary in fighting the Japa- 
nese in the jungle. Foxholes were more 
than six feet apart. The battalion laid no 
barbed wire or trip wire with hanging 
tin cans that rattled when struck by a 
man's foot or leg and warned of the ap- 
proach of the enemy. Thus, when dark- 
ness fell and the Japanese began their 
night harassing tactics— moving around, 
shouting, and occasionally firing— the 
imaginations of the tired and inexperi- 
enced American soldiers began to work. 
They thought the Japanese were all 
around them, infiltrating their perimeter 
with ease. One soldier reported that Jap- 
anese troops approached I Company, 
calling, in English, the code names of 
the companies of the 3d Battalion, such 
stereotypes as "come out and fight," and 
references to the Louisiana maneuvers. 12 

The men of the battalion, which had 
landed in the Russells the previous 
March, must have been familiar with the 
sights and sounds of a jungle night, but 
affected by weariness and the presence 
of the enemy, they apparently forgot. In 

15 169th Inf Hist, 20 Jun— 30 Sep 43, p. 5. 



their minds, the phosphorescence of 
rotten logs became Japanese signals. The 
smell of the jungle became poison gas; 
some men reported that the Japanese 
were using a gas which when inhaled 
caused men to jump up in their foxholes. 
The slithering of the many land crabs 
was interpreted as the sound of ap- 
proaching Japanese. Men of the 169th are 
reported to have told each other that 
Japanese nocturnal raiders wore long 
black robes, and that some came with 
hooks and ropes to drag Americans from 
their foxholes. In consequence the men 
of the battalion spent their nights ner- 
vously and sleeplessly, and apparently 
violated orders by shooting indiscrimi- 
nately at imaginary targets. 

Next day, the shaken 3d Battalion ad- 
vanced with I Company leading followed 
by L, M, Battalion Headquarters, and K 
Companies. It ran into machine gun fire 
from the Japanese trail block at 1055. I 
Company deployed astride Munda Trail, 
L Company maneuvered to the left, K 
w T as initially in reserve. M Company 
brought up its 81 -mm. mortars and heavy 
machine guns but could not use them 
profitably at first as banyan trees and 
undergrowth blocked shells and bullets. 
The mortar platoon then began clear- 
ing fields of fire by cutting down trees. 
B Company of the i^d also attacked the 
block from the south. 

I Company launched a series of frontal 
assaults but was beaten back by machine 
gun fire. Three platoon leaders were 
wounded in these attacks. K Company 
came out of reserve to deliver a frontal 
assault; its commander was soon killed. 
Neither it nor any of the other com- 
panies made progress. The Japanese were 
well dug in and camouflaged. Riflemen 

covered the automatic weapons. Fire 
lanes had been cut. The enemy weapons 
had little if any muzzle blast, and the 
Americans had trouble seeing targets. 
Some tried to grenade the enemy but 
were driven back before they could get 
close enough to throw accurately. At 
length the 81-mm. mortars got into 
action; observers operating thirty yards 
from the Japanese position brought 
down fire on it. Some Japanese are re- 
ported to have evacuated "Bloody Hill," 
as the Americans called it, that after- 
noon. At 1550 the 3d Battalion with- 
drew to dig in for the night. 13 After 
dark the Japanese harassed the 3d Bat- 
talion again. According to the 169th In- 
fantry, "a sleepless night was spent by all 
under continued harassing from enemy 
patrols speaking English, making horror 
noises, firing weapons, throwing hand 
grenades, swinging machetes and jump- 
ing into foxholes with knives." 14 

On 8 July, the 1st Battalion, 169th In- 
fantry, which had been behind the 3d 
within supporting distance, was ordered 
to bypass the 3d and move to the Barike 
while the 3d Battalion reduced the 
block. On 7 July General Wing had 
ordered Colonel Ross to use part of the 
i72d against the block, but apparently 
by the afternoon of 8 July no elements 
of the 17 2d except B Company had gone 
into action against it. On 8 July the 3d 
Battalion, 169th, and B Company, 17 2d, 
struck the block after a mortar prepara- 
tion and overran it. The 3d Battalion 

1,1 The 169th Infantry History (p. 4) claims that 
the block was destroyed on 7 July, and that a day 
was lost when the 1st Battalion, 169th, moved ahead 
of the 3d on 8 July. But messages in the 43d Di- 
vision G— 3 Journal indicate that the block was still 
active on 8 July. 

14 169th Inf Hist, p. 4. 



lost six men killed and thirty wounded, 
and suffered one case diagnosed as war 
neurosis, in reducing the block. The 
trail from Zanana to the Barike was open 
again, but the attack against Munda had 
been delayed by another full day. 

By late afternoon of 8 July, the ist 
Battalion, 169th, had reached the Barike 
River and made contact on its left with 
the 3d Battalion, i f j2&; A Company, 
169th, had been returned to its parent 
regiment; the 3d Battalion, i6gth, was 
behind and to the right of the 1st Bat- 
talion. With the two regiments on the 
line of departure, Hester and Wing were 
ready to start the attack toward Munda 
early on 9 July. Hester told Wing: "I 
wish you success." 15 

The Approach to the Main Defenses 

By 7 July General Hester, after con- 
ferences with General Wing and Colo- 
nels Ross and Eason, had abandoned 
the idea of the amphibious assault 
against Munda by the 3d Battalion, 103d 
Infantry, and the gth Marine Defense 
Battalion's tank platoon. He was prob- 
ably influenced in his decision by the 
strength of the Munda shore defenses. 
The plan for the attack on 9 July called 
for the 169th and 17 2d Regiments to ad- 
vance from the Barike, seize the high 
ground southwest of the river, and cap- 
ture the airfield. On the high ground— 
a complex of ridges that ran from Ilan- 
gana on the beach inland in a north- 
westerly direction for about three thou- 
sand yards— were the main Japanese de- 
fenses. The 173d Infantry was to move 
out astride the Munda Trail with the 
1st and 3d Battalions abreast. Each bat- 

» 43d Div G-3 Jnl, 8 Jul 43. 

talion zone would be three hundred 
yards wide. Battalions would advance in 
column of companies; each rifle company 
would put two platoons in line. The 
169th Infantry, maintaining contact on 
its left with the 3d Battalion, i72d, 
would advance echeloned to the right 
rear to protect the divisional right flank. 
The 1st Battalion was to advance abreast 
of the i72d; the 3d Battalion would 
move to the right and rear of the 1st. 

The regimental commanders planned 
to advance by 200-yard bounds. After 
each bound, they intended to halt for 
five minutes, establish contact, and move 
out again. They hoped to gain from one 
to two thousand yards before 1600. 

The division reserve consisted of the 
2d Battalion, 169th, which was to ad- 
vance behind the assault units. Antitank 
companies from the two regiments, plus 
Marine antiaircraft artillerymen, were 
defending the Zanana beachhead. In Oc- 
cupation Force reserve, under Hester, 
was the 3d Battalion, 103d Infantry, on 
Rendova. H Hour for the attack was set 
for 0630. 

General Barker's artillery on the off- 
shore islands inaugurated the first major 
attack against Munda at 0500 of 9 July 
with a preparation directed against rear 
areas, lines of communication, and sus- 
pected bivouac areas and command 
posts. After thirty minutes, fire was 
shifted to suspected centers of resistance 
near the line of departure. In one hour 
the 105-mm. howitzers of the 103d and 
169th Field Artillery Battalions, the 
155-mm. howitzers of the 136th Field 
Artillery Battalion, and the 155-mm. 
guns of the gth Marine Defense Bat- 
talion fired over 5,800 rounds of high 
explosive. Starting at 05 1 2 , four destroy- 



Troops of the 172D Infantry Wading Across a Creek, on the Munda Trail. 

ers from Admiral Merrill's task force, 
standing offshore in the Solomon Sea, 
opened fire at the area in the immediate 
vicinity of the airfield in accordance with 
plans prepared in consultation with Gen- 
eral Barker. Naval authorities had orig- 
inally wanted to fire at targets close to 
the line of departure as well, but the 
43d Division, fearing that the direction 
of fire (northeast to east) might bring 
shells down on its own troops, rejected 
the proposal. 16 Between 0512 and 0608, 
the destroyers fired 2,344 5-inch rounds. 
At 0608, four minutes before the bom- 
bardment was scheduled to end, some 
Japanese planes dropped bombs and 
strafed one ship; the destroyers retired. 

M Merrill thought the 43d Division was generally 
too cautious. See Morison, Breaking the Bismarcks 
Barrier, p. 179. 

Then Allied planes from Guadalcanal 
and the Russells took over. Fifty-two 
torpedo bombers and thirty-six dive 
bombers dropped seventy tons of high 
explosive bombs and fragmentation clus- 
ters on Munda. Now it was the infan- 
try's turn. 

H Hour, 0630, came and went, but 
not a great deal happened. The ist Bat- 
talion, 169th Infantry, reported that it 
was ready to move but could not under- 
stand why the i72d Infantry had not ad- 
vanced. At 0930, General Wing was in- 
formed that no unit had yet crossed the 
line of departure. Several factors seem 
to have caused the delay. Movement as 
usual was an ordeal. The Bar ike was 
flooded. Soldiers, weighted with weap- 
ons, ammunition, and packs, had to wade 
through waist-to-shoulder-deep water. 



The river, which had several tributaries, 
wound and twisted to the sea. It crossed 
the Munda Trail three times; the spaces 
between were swampy. The men, sweat- 
ing in the humid heat, struggled to keep 
their footing, and pulled their way along 
by grabbing at roots and undergrowth. 
Leading platoons had to cut the wrist- 
thick rattan vines. 

Although patrols of New Georgians, 
Fijians, Tongans, New Zealanders, and 
Americans had reconnoitered the area, 
their information could not always be 
put to good use. There was no accurate 
map on which to record data, nor were 
there any known landmarks. 

In the jungle, orthodox skirmish lines 
proved impractical. As men dispersed 
they could not be seen and their leaders 
lost control. At any rate, movement off 
the trails was so difficult that most units 
moved in columns of files, the whole 
unit bound to one trail. Thus one or 
two Japanese, by firing on the leading 
elements, could halt an entire battalion. 

The Occupation Force intelligence 
officer had estimated that the main Japa- 
nese defenses lay 1,600 yards from the 
Barike, anchored on Roviana Lagoon 
and extending inland to the northwest. 
This was correct, except that the defense 
line on the ridges was actually about 
2,500 yards from the Barike's mouth. 
Beyond the main defenses, the Japanese 
outposts, using rifles, machine guns, and 
sometimes mortars and grenade dis- 
chargers, were well able to delay the 

At 1030 General Barker returned to 
the 43d Division command post from a 
tour of the front and reported that at 
1000 the 17 2d Infantry was a hundred 
yards beyond the Barike, but that the 

169th was still east of the river. The 
only opposition had come from the out- 
post riflemen that the Americans usually 
called "snipers." At the time these were 
believed, probably erroneously, to be 
operating in the treetops. 17 

Japanese fighter aircraft appeared over 
New Georgia during the day; the Allied 
air power prevented any from getting 
close enough to strafe the attacking 

By 1630, when it dug in for the night, 
the 17 2d had gained some eleven hun- 
dred yards. 18 The 169th had made no 
progress to speak of. The 1st Battalion 
got one hundred yards west of the 
Barike; the other two apparently re- 
mained east of the river. 

The 169th was facing about the same 
obstacles as the i72d, but it is possible 
that the 169th was a badly shaken regi- 
ment before the attack began. 1 " The 
night before the attack, 8-9 July, the 3d 
Battalion was bivouacked near Bloody 
Hill, and the other two lay to the west. 
When the Japanese made their presence 
known to the three battalions, or when 
the Americans thought there were Japa- 
nese within their bivouacs, there was a 
great deal of confusion, shooting, and 

17 Whereas the Japanese, like the Allies, used trees 
whenever possible for observation posts, it is doubt- 
ful that "snipers" used many trees in the jungle. 
See Miller, Guadalcanal: The First Offensive, p. 
318. Anyone who has ever climbed a tree in the 
jungle can testify to the difficulties a man with a 
rifle would encounter— lack of visibility, tree limbs 
in the way, and the innumerable little red ants 
whose bite is like the prick of needles. 

18 From 1100, 8 July, to 1300, 9 July, this regi- 
ment was commanded by Lt. Col. Charles W. Cap- 
ron. Colonel Ross, wounded on 30 June, had ap- 
parently been ordered to Rendova for medical 

* The i72d was either not subjected to night 
harassing or was not sufficiently bothered by it to 
report it. 



stabbing. Some men knifed each other. 
Men threw grenades blindly in the dark. 
Some of the grenades hit trees, bounced 
back, and exploded among the Ameri- 
cans. Some soldiers fired round after 
round to little avail. In the morning no 
trace remained of Japanese dead or 
wounded. But there were American 
casualties; some had been stabbed to 
death, some wounded by knives. Many 
suffered grenade fragment wounds, and 
50 percent of these were caused by frag- 
ments from American grenades. These 
were the men who had been harassed by 
Japanese nocturnal tactics on the two 
preceding nights, and there now ap- 
peared the first large number of cases 

diagnosed as neuroses. The regiment was 
to surfer seven hundred by 31 July. 

The 43d Division resumed the attack 
on 10 July. The 173d Infantry, report- 
ing only light opposition, advanced a 
considerable distance. The 169th In- 
fantry, with the 1st Battalion in the lead 
and the 2d Battalion to its right rear, ad- 
vanced successfully until it reached the 
point where the Munda Trail was inter- 
sected by a trail which ran southeast to 
the beach, then circled to the southwest 
to the native villages of Laiana and Ilan- 
gana. Reaching this junction about 1330 
after crossing a small creek on two felled 
tree trunks, the leading battalion was 
halted by machine gun fire. This fire 



came from rising ground dominating 
the trail junction, where Capt. Bunzo 
Kojima, commanding the gth Company, 
229th Infantry, had established a camou- 
flaged trail block. He employed one rifle 
platoon, reinforced by a machine gun 
section, some 90-mm. mortars, and ele- 
ments of a 75-mm. mountain artillery 
battalion. When the 1st Battalion was 
stopped, Colonel Eason decided to blast 
the strong point. While the infantry 
pulled back a hundred yards, the i6{)th's 
mortars and the Occupation Force artil- 
lery opened fire. Barker's guns fired over 
four thousand rounds of 105-mm. and 
155-mm. high explosive, shattering trees, 
stripping the vegetation, and digging 
craters. 20 Coincident with this bombard- 
ment, eighty-six Allied bombers (SBD's 
and TBF's) unloaded sixty-seven tons of 
bombs on Lambeti Plantation and 
Munda. During the artillery bombard- 
ment Kojima's men lay quiet but when 
the fire ceased they immediately stood 
to their guns and halted the American 
infantrymen when they attacked. At the 
day's end, the Japanese were still on the 
high ground; the 169th Infantry, after 
advancing about fifteen hundred yards, 
was forced to bivouac in a low swampy 
area. The American commanders con- 
cluded that they were nearing a main 
defensive line. They were right. The 
high ground to their front contained the 
main Japanese defenses that were to re- 
sist them for weeks. 

Laiana Beachhead 

By 11 July the advancing regiments 
were still in trouble. Progress had been 

20 One fortunate concomitant of artillery fire was 
better visibility as the heavy shellings tore the 
jungle apart. 

slowed by the enemy, and also by the 
supply problems arising from the fact 
that the troops had landed five miles 
east of their objective and thus com- 
mitted themselves to a long march 
through heavy jungle. Now the regi- 
ments, in spite of their slow advance, had 
outdistanced their overextended supply 

The 118th Engineer Battalion had 
made good progress in building a jeep 
trail from Zanana to the Barike River. 
Using data obtained from native scouts, 
the engineers had built their trail over 
high, dry ground, averaging one half to 
three quarters of a mile per day. There 
was little need for corduroying with logs, 
a time-consuming process. When they 
ran into trees too big to knock down 
with their light D-3 bulldozers, the en- 
gineers blasted them with dynamite. 
Lacking heavy road-graders, the 118th 
could not make a two-lane, amply 
ditched road, but it managed to clear a 
one-lane track widened at regular inter- 
vals to permit two-way traffic. Near a 
five-foot-deep, fast-running stream east of 
the Barike the engineers hit soft mud. 
To get to ground firm enough to permit 
construction of footbridges and two 
thirty-foot trestle bridges, they were 
forced to swing the road northward par- 
allel to the river for two and one-half 
miles to get to a firm crossing. The ad- 
vancing regiments crossed the Barike on 
9 July, but several days were to elapse 
before the bridges were completed. 

Thus there was a gap between the end 
of the road and the front. To bridge the 
gap, nearly half the combat troops were 
required to carry forward ammunition, 
food, water, and other supplies, and to 
evacuate casualties. Allied cargo planes 



were used to parachute supplies to the 
infantry, but there were never enough 
planes to keep the troops properly sup- 

With fighting strength reduced by the 
necessity for hand carry, with his right 
flank virtually exposed, and his extended 
supply line open to harassment by the 
enemy, Hester decided, on 10 July, to 
change his plan of attack in order to 
shorten the supply line. If a new beach- 
head could be established at Laiana (a 
native village about two miles east by 
south from Munda airfield), some five 
thousand yards would be cut off the sup- 
ply line. Patrols, operating overland and 
in canoes, examined Laiana beach at 
night and reported that it was narrow 
but suitable, with a coral base under the 
sand. Unfavorable aspects included a 
mangrove swamp back of the beach and 
the fact that the Japanese main defenses 
appeared to start at Ilangana, only five 
hundred yards southwest of Laiana, and 
arch northwest toward the Munda Trail. 

But the advantages outweighed the 
disadvantages. Hester ordered the l'jzd 
Infantry to swing southward to Laiana, 
seize and hold a beachhead from the 
land side, then advance on Munda. The 
i6gth Infantry was to continue its at- 
tempt to drive along the Munda Trail. 
Hester ordered the reinforced 3d Bat- 
talion, 103d Infantry, at Rendova, to be 
prepared to land at Laiana after the 
173d had arrived. 

At 1000, 11 July, the l^sd Infantry 
disengaged from the attack, turned 
south, and started moving toward shore 
through knee-deep mud. The regiment 
tried to keep its move a secret, but Japa- 
nese patrols quickly observed it, and 

Jeep Trail From Zanana, built 
through heavy jungle by 118th Engi- 
neer Battalion, 13 July 1943- 

mortar fire soon began hitting it. The 
wounded were carried along with the 
regiment. The advance was halted about 
midafternoon after a gain of some 450 
yards. Both 1st and 3d Battalions (the 
2d had remained behind to block the 
trail and thus cover the rear until the 
169th could come up) reported running 
into pillboxes. Aside from the mortar 
shelling and some infiltration by patrols 
between the 17 2d and the 169th, the 
Japanese appeared to have stayed fairly 

The march was resumed on 12 July 
with the hope of reaching Laiana be- 
fore dark, for the regiment had not re- 



ceived any supplies for two days. Colonel 
Ross reported that the carrying parties 
equaled the strength of three and one- 
half rifle companies. Despite this fact, 
and although food and water were ex- 
hausted, the regiment kept moving until 
late afternoon when leading elements 
were within five hundred yards of 
Laiana. There machine gun and mortar 
fire halted the advance. At this time 
scouts confirmed the existence of pill- 
boxes, connected by trenches, extending 
northwest from Ilangana. The pillboxes, 
which the Americans feared might be 
made of concrete, housed heavy machine 
guns, and were supported by light ma- 
chine guns and mortars. 

That night (12-13 July) Japanese 
mortars registered on the ljzd's bivouac, 
and the troops could hear the Japanese 
felling trees, presumably to clear fields 
of fire. 

His hungry, thirsty regiment was with- 
out a line of communications, and Col- 
onel Ross, concerned over the Japanese 
patrols in his rear, had to get to Laiana 
on 13 July. With the artillery putting 
fire ahead, the 17 2d started out through 
mangrove swamp on the last five hun- 
dred yards to Laiana. The enemy fire 
continued. The advance was slow, but 
late afternoon found the 173d in posses- 
sion of Laiana. It organized the area for 
defense while patrols sought out the 



LCM's Approaching Laiana, New Georgia, under Japanese artillery fire, 14 
July 1943. The Tank Platoon of the 9th Marine Defense Battalion is aboard 
these landing craft. 

Japanese line to the west. That night 
twelve landing craft left Rendova to 
carry food and water to Laiana and 
evacuate the wounded. For some reason 
the 17 2d failed to display any signals. 
The landing craft, unable to find the 
right beach, returned to Rendova. 

When the i72d was nearing Laiana 
on 13 July, General Hester ordered the 
3d Battalion, 103d Infantry, 43d Divi- 
sion, to be prepared to land at 0900 the 
following morning. The Tank Platoon 
of the 9th Marine Defense Battalion was 
attached; to help the tanks and to aid in 
the reduction of fixed positions, en- 
gineers (bridge builders, demolitions 

men, flame thrower operators, and mine 
detector men) were also attached. 

The reinforced battalion, loaded in 
LCP(R)'s and LCM's, rendezvoused at 
daybreak of 14 July in Blanche Channel. 
When the daily fighter cover arrived 
from the Russells, the landing craft 
started for Laiana. With the 17 2d already 
holding the beachhead, the first wave 
landed peacefully at 0900. Reefs forced 
some craft to ground in waist-deep water, 
but the hungry soldiers of the 173d 
helped unload them. As the LCM's 
neared shore Japanese artillery shells be- 
gan falling on the water route and on 
the landing beach. To blind the Japa- 



nese observers, the field artillery fired 
more than five hundred white phos- 
phorous rounds as well as high explosive 
at suspected Japanese gun positions and 
observation posts on Munda Point and 
on the high ground (Bibilo Hill) north- 
east of Munda field. The Japanese artil- 
lery did no damage. 

General Sasaki reported that he had 
repulsed the landing, and that the Amer- 
icans had lost, of seventy landing craft, 
thirteen sunk and twenty damaged. 
Nevertheless, 8th Area Army headquart- 
ers appears to have learned that the 
landing had succeeded. 

Once ashore, 43d Division engineers 
began building a jeep trail from Laiana 
north to the 169th Infantry. Supplies 
came in for the 17 2d, and its wounded 
men were evacuated. Telephone crews 
laid an underwater cable between 
Zanana, Laiana, and General Barker's 
artillery fire direction center. 

The 3d Battalion, 103d, was still in 
division reserve, but Colonel Ross was 
authorized to use it in case of dire need. 
He committed L Company to fill a gap 
between the 2d and 3d Battalions of the 
173d on the morning of 15 July when 
the 173d was making an unsuccessful 
attack toward Ilangana. Soldiers of the 
Antitank Platoon of the 3d Battalion, 
103d, disassembled a 37-mm. gun, car- 
ried it forward, reassembled it on the 
front line, and destroyed three pillboxes 
with direct fire. This was the only suc- 
cess; the day's end found the i72d still 
facing the main enemy defense line. 

The Seizure of Reincke Ridge 

While the 17 2d had been driving to 
Laiana and getting ready to attack west- 
ward, the 169th Infantry was pushing 

against the high ground to the north. On 

10 July, the day before the i72d turned 
southward, the 169th had been halted. 
It faced Japanese positions on the high 
ground which dominated the Munda- 
Lambeti trail junction. The Munda 
Trail at this point led up to a draw, 
with hills to the north (right) and south 
(left). The Japanese held the draw and 
the hills. 

The regiment renewed the attack on 

11 July just before General Hester re- 
placed Colonel Eason with Col. Temple 
G. Holland, but made no gains. When 
Holland took over the regiment, he 
ordered the advance postponed until the 
next morning. The exact nature of the 
Japanese defenses was not yet completely 
clear, but it was evident that the Japa- 
nese had built mutually supporting pill- 
boxes on the hills. 

Holland's plan for 12 July called for 
the 1st Battalion to deliver the attack 
from its present position while the 2d 
Battalion enveloped the Japanese left 
(north) flank. 21 The 3d Battalion, tem- 
porarily in division reserve, would be re- 
leased to the regiment when the trail 
junction was secure. The 169th attacked 
as ordered but bogged down at once, 
partly because it became intermingled 
with elements of the 17 2d, which was 
starting for Laiana. When the units were 
disentangled the two battalions attacked 
again. The 1st Battalion ran head on into 
Japanese opposition but reported a gain 
of three hundred yards. The 2d Battalion 
received enfilading fire from the north- 
ernmost ridge but kept its position. A 

"Ltr, Col Holland to Gen Smith, Chief of Mil 
Hist, 12 Oct 53, no sub, OCMH. 



second attack, supported by a rolling 
barrage, was attempted in the afternoon. 
The infantry, unable to keep pace with 
the barrage which moved forward at the 
rate of ten yards a minute, fell behind 
and halted. At the day's end, Holland, 
who reported to Hester that his regi- 
ment was badly disorganized, asked Gen- 
eral Mulcahy for air support the follow- 
ing day. 

Next morning, 13 July, after thirty 
minutes of artillery fire and a twelve- 
plane dive-bombing attack against the 
south ridge, the 169th attacked again. 
All three battalions were committed. 
The 2d Battalion, in the center, was to 
assault frontally up the draw while the 
1st Battalion, on the right, and the 3d 
Battalion on the left, moved against the 
north and south ridges with orders to 
envelop the Japanese. 

The 3d Battalion, with I and L Com- 
panies in line and M in support, strug- 
gled forward for four hours. 22 It pushed 
four hundred to five hundred yards into 
the Japanese lines and managed to secure 
its objective, the south ridge, which it 
named Reincke Ridge for Lt. Col. 
Frederick D. Reincke, who had replaced 
Stebbins in command on 8 July. 

The other two battalions were not as 
successful. The 2d Battalion, with E and 
F Companies in line and G in support, 
met machine gun fire in the draw, 
halted, was hit by what it believed to be 
American artillery fire, and pulled back. 
The 1st Battalion, attacking the north 
ridge, found it obstructed by fallen 
limbs from blasted trees and by shell 

" K Company had been detached to guard the 
regimental command post. 

and bomb craters. The Japanese who 
had survived the bombardments opened 
fire from their pillboxes and halted 
the assaulting companies. The battalion, 
now operating without artillery or mor- 
tar support, tried to assault with rifle 
and bayonet. Some men started to climb 
to the ridge crest but were killed or 
wounded by machine gun fire. B Com- 
pany lost three of its four officers in the 
attempt. Japanese artillery and mortar 
fire cut communication to the rear. The 
battalion returned to its original posi- 

The 1st and 2d Battalions took posi- 
tions on the flanks and rear of the 3d 
Battalion, which held Reincke Ridge. 
The Japanese held the north ridge and 
the draw. To the west they held the 
higher ground called Horseshoe Hill. To 
the south was the gap left by the i72d 
when it turned south. In spite of the 3d 
Battalion's exposed situation Holland 
and Reincke decided to hold the hard- 
won position which was the only high 
ground the i6gth possessed. Its posses- 
sion was obviously vital to the success 
of an attack against the main enemy de- 

All that night and all the next day (14 
July) the Japanese tried to push the 3d 
Battalion from Reincke Ridge. I Com- 
pany was hit hard but held its ground 
with the loss of two men killed and nine- 
teen wounded. Artillery and mortar 
shells kept exploding on the ridge top, 
while Japanese machine guns covered 
the supply route to the rear. During its 
first twenty-four hours on the ridge, 
Reincke's battalion suffered 101 casual- 
ties; L Company consisted of just fifty- 
one enlisted men by the end of 14 July. 



During part of the time no medical 
officer was present, but the battalion 
medical section under S/Sgt. Louis Gul- 
litti carried on its duties of first aid and 

On the same day Holland reorganized 
the other two battalions. The regimental 
Antitank Company had landed at Zanana 
on 13 July and been assigned the task 
of carrying supplies forward from the 
trail's end. This task had eased, because 
the engineers finished bridging the 
Barike on 12 July and by 14 July had 
extended the trail to within five hun- 
dred yards of the lGgth's front lines. 
Rations, water, and ammunition were 
parachuted to the regiment on 14 July. 
Colonel Holland relieved part of the 
Antitank Company of its supply duties 
and assigned sixty of its men to the 2d 
Battalion, twenty to the 1st. He also sent 
patrols south to cover the gap to his left. 
Late in the afternoon he reported to 
Hester that morale in his regiment had 

Next day the 1st Battalion, 145th In- 
fantry, landed at Zanana and was im- 
mediately attached to the 43d Division 
with orders to advance west and relieve 
part of the 169th on the line. The bat- 
talion reached the regiment at 1700. 
Colonel Holland put it in regimental 
reserve pending the completion of the 
operations against the hills in front of 

Operations against Munda airfield 
had gone very slowly but by 15 July had 
achieved some success. Liversedge had 
captured Enogai and while waiting for 
another battalion was getting ready to 
attack Bairoko. The i6gth Infantry had 
some high ground and was in contact 
with the main enemy defense line. The 

17 2d Infantry was also in contact with 
the main Japanese defenses, and the new 
beachhead at Laiana would soon shorten 
the supply line. 


While Hester's men had been attain- 
ing limited tactical successes, unusual 
medical problems had appeared within 
his division. 

Enemy resistance was not great at 
first. Some go men of the 43d Division 
were killed up to 17 July; 636 were 
wounded. Other men had been injured 
by vehicle collisions, falling trees, acci- 
dental explosions, and the like. Disease 
had put over 1,000 men out of action. 23 

Diarrhea and dysentery, ailments 
helped along by improper field sanita- 
tion, were prevalent in early July. They 
put men on the sick list for several days. 
Skin fungus infected about one quarter 
of the men. And there was always ma- 
laria. Although malaria control measures 
seem to have been carried out so con- 
scientiously that few new cases broke out 
in the Occupation Force, all the troops 
had been in the Solomons for some time 
and there were always recurrent cases. 

An especially large number of casual- 
ties was caused not by wounds or infec- 
tious disease but by mental disturbance. 
Between fifty and a hundred men were 
leaving the line every day with troubles 
which were diagnosed as "war neuroses." 
Col. Franklin T. Hallam, surgeon of the 
XIV Corps, arrived in New Georgia on 
14 July when mental troubles were at 
their height. In Hallam's opinion, "war 
neurosis" was a "misnomer in most in- 
stances," because men suffering simply 

XIV Corps G-3 Jnl, 17 Jul 43. 



from physical exhaustion "were erro- 
neously directed or gravitated through 
medical channels along with the true 
psychoneurotics and those suffering with 
a temporary mental disturbance cur- 
rently termed 'WAR NEUROSIS.' " 24 

These unfortunate men "who had not 
changed clothes or had two continuous 
hours of sleep all had the same expres- 
sion. Their hair was matted and muddy, 
and beards were \/ 2 mcn m length, eyes 
were sunk in, dark, and had a strained 
expression. Gait was plodding and 
methodical, no spring or bounce. When 
they stopped walking they fell in their 
tracks, until it was time to proceed 
again." 25 Colonel Hallam's description 
is even more graphic: 

At least 50% of these individuals requir- 
ing medical attention or entering medical 
installations were the picture of utter ex- 
haustion, face expressionless, knees sagging, 
body bent forward, arms slightly flexed and 
hanging loosely, hands with palms slightly 
cupped, marked coarse tremor of fingers 
. . ., feet dragging, and an over-all appear- 
ance of apathy and physical exhaustion. 
About 20% of the total group were/highly 
excited, crying, wringing their hands, mum- 
bling incoherently, an expression of utter 
fright or fear, trembling all over, startled at 
the least sound or unusual commotion, 
having the appearance of trying to escape 
impending disaster. Another 15% showed 
manifestations of the various types of true 
psychoneurotic complexes. The remaining 
15% included the anxiety states, and those 
with various bizarre somatic disturbances. 
These were the individuals whose symptoms 

"Ltr, Col Hallam to The Surgeon, USAFISPA, 
31 Oct 43, sub: Med Service, New Georgia Cam- 
paign, p. 31. 

" This description comes from a personal ac- 
count, "Medic on Munda," by Capt, Joseph Risman 
(Medical Corps), 169th Infantry, and is quoted in 
SOPACBACOM, History of the New Georgia Cam- 
paign, Vol. I, Ch III, p. 26, OCMH. 

were of insidious onset, starting with in- 
somnia, vague digestive symptoms, bad 
dreams, frequency of urination, irritability, 
diminished ability to concentrate, and a 
generally reduced efficiency in the per- 
formance of assigned duties. 26 

Of about 2,500 men in the New 
Georgia Occupation Force whose trou- 
bles were diagnosed as "war neuroses" 
between 30 June and 30 September, 
the 43d Division contributed 62 per- 
cent during the period 30 June-3 1 July. 
About 1,500 cases came from the three 
infantry regiments of the 43d Division: 
700 from the 169th Infantry, 450 from 
the i^2d Infantry, and 350 from the 
103d Infantry. 27 

Attempting to explain this mental 
trouble, Hallam divided the causes into 
two groups he termed "basic causative 
factors" and "precipitating causative fac- 
tors." Basic causes involved leadership, 
orientation, discipline, and physical fit- 
ness. Units with poor leaders were more 
apt to have trouble than those in which 
the standard of leadership was high. In 
some units there was a direct correlation 
between the incidence of mental trou- 
bles among the leaders and among the 
led. When soldiers were not adequately 
oriented— not told what was going on, 
what their objectives were, and what they 
were expected to do— they were more 
apt to become excited by loose talk and 

J "Ltr, Hallam to The Surgeon, USAFISPA, 31 
Oct 43, sub: Med Service, New Georgia Campaign, 
PP- 36-37- 

"Ibid., p. 32. Of the 2,500 cases occurring from 
30 June to 30 September, the 43d Division had 
about 1,950 or 79 percent; the 37th Division, 200 
or 8 percent; the 25th Division, 150 or 6 percent; 
Navy and Marine Corps units, 200 or 8 percent. In 
July, the New Georgia Occupation Force had 1,750 
cases or 70 percent, in August, 650 or a6 percent, 
and in September, 100 or 4 percent. 



wild rumors. The significance of lack of 
proper discipline and physical fitness in 
any military organization, but especially 
in one engaged in battle, is perfectly ob- 
vious. Interestingly enough, however, 
Hallam noted that men "with borderline 
physical defects, consisting principally of 
eye, teeth, joint, weight, and feet de- 
fects, did not break, but did some of the 
best fighting." 28 Remarkably few men 
wounded in action became neurosis 
cases, perhaps because their knowledge 
that they would be evacuated eased their 
mental strain. 

The basic causes, of course, were pres- 
ent in some units when they came to 
New Georgia. It was Hallam's opinion 
that men affected by any of the basic 
causes were triggered into mental dis- 
turbance by the precipitating factors, 
which were combat fatigue, enemy ac- 
tion, noise, and mass hysteria. Combat 
fatigue, the almost unutterable physical 
and mental weariness that comes from 
long stress and strain in battle, probably 
accounted for half the diagnoses of war 
neuroses. The most effective enemy ac- 
tion was the kind which so seriously dis- 
turbed the 169th Infantry— the real, and 
occasionally the wholly imaginary, noc- 
turnal harassing tactics of the Japanese. 
Although aerial bombardment was also 
effective, the noises to which Hallam re- 
ferred were not the sounds of guns fir- 
ing and shells bursting, but the natural 
sounds of a jungle night, the breezes, 
branches, birds, and land crabs. These 
caused great anxiety among men to 
whom they were unfamiliar. On oc- 
casion mass hysteria took over; mental 
breaks spread like infection among 

a Ibid., p. 35. 

Most of the mental cases, and espe- 
cially those caused by fatigue, Hallam 
believed, could have been cured by a 
few days in a rest camp in the combat 
area. Sedatives, sleep, clean clothes, 
baths, shaves, good food, relief from 
duties, and recreation would soon have 
enabled the men to return to their units. 
But up to mid-July there were no rest 
camps, nor even any real hospital fa- 
cilities, in New Georgia. The 43d Di- 
vision, about 30-35 percent under- 
strength in medical officers and enlisted 
men, had only a 125-bred clearing station 
to care for casualties. 29 Men requiring 
more than twenty-four hours of medical 
treatment were being evacuated, usually 
by water, to Guadalcanal, with the re- 
sult that casualties frequently did not 
reach hospitals until three days after 
they had been taken out of the line. 

These medical problems, coupled with 
the slow progress of ground operations 
up to mid-July, caused serious concern 
to all the responsible higher command- 

Command and Reinforcements 

As early as 10 July, Generals Hester 
and Wing were far from pleased with the 
performance of all units and command- 
ers. On 10 July Wing, who had visited 
the command post of the 3d Battalion, 
169th Infantry, on 8 July, directly 
ordered the regimental commander to 
relieve the 3d Battalion's commander 
and put Colonel Reincke in his place. 

Three days prior to this relief, the 
145th Infantry Regiment (less the 3d 
Battalion, serving under Liversedge) of 
the 37th Division, which had been stand- 

w The 17th Field Hospital came up from Guadal- 
canal and opened on New Georgia on 28 July. 



ing by on Guadalcanal in area reserve, 
had been dispatched to Rendova. The 
first echelon sailed on 7 July, the second 
two days later. The regimental com- 
mander, Colonel Holland, had hardly 
arrived on Rendova when Hester re- 
lieved the commander of the 169th In- 
fantry and ordered Colonel Holland to 
take over the regiment temporarily. Also 
relieved were the executive, intelligence, 
and operations officers of the 169th. 
Leaving Lt, Col. Theodore L. Parker in 
command of his old regiment, Holland 
took his own executive, intelligence, and 
operations officers and eighteen enlisted 
men from the 145th to headquarters of 
the 169th. 

Meanwhile problems of higher com- 
mand for New Georgia had not ceased 
to concern Admirals Halsey and Turner 
and especially General Harmon. On 5 
July Harmon was on Guadalcanal, as 
were Turner and General Griswold. 
After informing Turner and Griswold 
of his views, he radioed to Halsey a rec- 
ommendation that the forward echelon 
of the XIV Corps staff be sent to New 
Georgia about 8 July to prepare, under 
Hester, to take over supply, administra- 
tion, and planning. Once Munda airfield 
fell, Harmon urged, Griswold should 
become commander of the New Georgia 
Occupation Force. This would free 
Hester to reorganize his main striking 
force and directly command the attack 
against Vila in Kolombangara. Such a 
change was necessary, Harmon ex- 
plained, because Hester's small staff was 
not capable of bearing the responsibil- 
ities that would soon be thrust on it. 30 

^Rad, Harmon to COMSOPAC, 5 Jul 43, Hq 
SOPACBACOM File. Suppl New Georgia Material, 

Admiral Turner was not a man given to 
avoiding responsibility or yielding au- 
thority. Harmon wrote later, in explain- 
ing his reasons for urging a change in 
command, that Turner was "inclined 
more and more to take active control of 
land operations." 31 In his message to 
Halsey, he did not make this point. The 
South Pacific commander replied to Har- 
mon the next day, telling him to aug- 
ment Hester's 43d Division staff as he 
saw fit. Halsey wished to discuss with 
Harmon the recommendations on super- 
seding Hester before reaching a decision. 
On the same day Halsey directed Turner 
to prepare plans for Kolombangara in 
consultation with Hester. 32 

The next day the irascible Turner 
presented his views to Halsey in very 
mild terms. Expressing regret over the 
necessity for disagreeing with Harmon, 
he strongly urged that Hester retain 
command of the New Georgia Occupa- 
tion Force. Griswold and his staff were 
excellent, Turner agreed, but Hester 
was conducting operations "in a manner 
much to be admired." Superseding him 
would hamper the operation "by induc- 
ing a severe blow to morale." 33 

At this point Harmon, a peppery, wiry 
man, grew impatient. He boarded his 
B-17 and flew to Halsey in Noumea. 
". . . before nightfall," he later related, 

31 Ltr, Harmon to Handy, 15 Jul 43, quoted in 
part in Hq SOPACBACOM File, Suppl New 
Georgia Material, OCMH. 

2a Rad, COMSOPAC to Harmon, 6 Jul 43, Hq 
SOPACBACOM File, Suppl New Georgia Material, 
OCMH; COMSOPAC War Diary, 6 Jul 43. 

^Rad, CTF 31 to COMSOPAC, 7 Jul 43, Hq 
SOPACBACOM File, Suppl New Georgia Material, 



"Admiral Halsey approved the course of 
procedure I had recommended." 31 

Griswold received instructions on 10 
July to take six officers from his staff and 
fly to New Georgia on 1 1 July in an 
amphibian plane. The remainder of the 
XIV Corps staff would follow by water 
on 12 July. On orders from Halsey, 
which the admiral expected to issue after 
the capture of Munda airfield, Griswold 
would assume command of the New 
Georgia Occupation Force. Turner's au- 
thority over the Occupation Force would 
cease, but he was to continue to support 
the operation. Halsey repeated to Turner 
his instructions regarding plans for 
taking Kolombangara, and told him 
that, if Griswold approved the idea, 
Hester would command the ground 
forces in the attack. 35 

Griswold arrived at Rendova on l i 
July just as Hester and Wing were 
changing their plan of attack against 
Munda and sending the i72d Infantry 
to seize the Laiana beachhead. The XIV 
Corps commander was not long in reach- 
ing a judgment regarding operations to 

General Harmon, at his headquarters 
in Noumea, wrote an optimistic letter 
to Washington on the morning of 13 
July. He reported that operations in 
New Georgia seemed to be progressing 
favorably. He did not send the letter, for 
later in the morning he received a radio- 

^Ltr, Harmon to Handy, 15 Jul 43, quoted in 
part in Hq SOPACBACOM File, Suppl New 
Georgia Material, OCMH; COMSOPAC War Diary, 
9 Jul 43- 

30 Rad, COMSOPAC to CTF 31, 9 Jul 43; and 
Rad, COMGENSOPAC to CG XIV Corps, 10 Jul 
43, in Hq SOPACBACOM File, Suppl New Georgia 
Material, OCMH; XIV Corps G-3 Jnl, 10-11 Jul 

gram from General Griswold, who said, 
"From an observer point of view things 
are going badly." Griswold urged that 
the 25 th Division and the remainder of 
the 37th Division be sent into the battle 
at once. Although he reported, "Enemy 
resistance to date not great," he did not 
think the 43d Division would ever take 
Munda. It was, he declared, "about to 
fold up." 36 

This message had an immediate effect. 
Halsey met w r ith Harmon and informally 
appointed him as his deputy. He ordered 
Harmon to "assume full charge of and 
responsibility for ground operations in 
New Georgia," and "to take whatever 
steps were deemed necessary to facilitate 
the capture of the airfield." 37 

Before leaving for Koli Point on 
Guadalcanal to be nearer the scene of 
action, Harmon ordered Griswold to 
hasten his preparations for assuming 
command on New Georgia. All ground 
forces assigned for the operation, he told 
Griswold, would be available by the 
time he assumed command. Harmon 
promised to alert one regimental com- 
bat team of the veteran 25th Division 
for movement, but it would be dis- 
patched to New Georgia only if he 
specifically approved. 

Of the assigned 37th Division forces, 
the 145th Infantry, like the 136th Field 
Artillery Battalion, was already on hand 
in New Georgia, the ist and 2d Battal- 
ions at Rendova and the 3d Battalion un- 
der Liversedge along with 3d Battalion, 
148th Infantry. Admiral Turner at once 

88 Rad, Griswold to Harmon, 13 Jul 43, quoted 
in SOPACBACOM, History of the New Georgia 
Campaign, Vol. I, Ch. Ill, p. 39, OCMH. 

a ' Harmon, The Army in the South Pacific, p. 8; 
Halsey, Narrative Account of the South Pacific 
Campaign, p. 7. Both in OCMH. 



Rear Adm. Theodore S. Wilkinson (left) and Lt. Gen. Millard F. Harmon in 
the chart room of the transport McCawley. 

ordered Col. Stuart A. Baxter, com- 
manding the 148th Infantry in the Rus- 
sell Islands, to alert Headquarters, the 
1st and 2d Battalions, and the Antitank 
Company of his regiment for immediate 
movement to New Georgia. These move- 
ments would put two full infantry regi- 
ments of the 37th Division in New 

On the 16th, Griswold proposed that 
the 37th Division units operate under 
control of their division commander, 
Maj. Gen. Robert S. Beightler, and that 
Beightler and his senior staff officers fly 

to New Georgia for conferences and per- 
sonal reconnaissance. Harmon agreed, 
and Beightler left for New Georgia in a 
PBY on 19 July. 

On arriving at Guadalcanal, Harmon 
ordered Maj. Gen. J. Lawton Collins, 
commanding the 25th Division, to get 
one regimental combat team ready for 
transfer to New Georgia. Collins, who 
on Griswold's departure had become 
island commander and as such respon- 
sible for Guadalcanal's defense, decided 
that the 161st Regimental Combat Team 
could most easily be spared from its de- 



fense missions. On 14 July he directed 
Col. James L. Dalton, II, regimental and 
combat team commander, to be ready to 
move on twelve hours' notice. 38 

The next day Admiral Turner was re- 
lieved of his posts of Commander, South 
Pacific Amphibious Force (III Amphibi- 
ous Force and Task Force 32), and Com- 
mander, New Georgia Attack Force 
(Task Force 31). This relief apparently 
had nothing to do with recent events on 
New Georgia. Admiral Nimitz, then pre- 
paring for the great Central Pacific drive 
that was to start with the invasion of the 
Gilberts in November 1943, had directed 
Halsey to send Turner to Hawaii. 
Turner departed on the 15th, and dur- 
ing the next two years commanded the 
V Amphibious Force in the invasions of 
the Gilberts, the Marshalls, the Marianas, 
Iwo Jima, and Okinawa. His posts in 
the South Pacific were taken over by Ad- 
miral Wilkinson, until then Halsey's 
deputy commander. 

On the day Turner left, Harmon 
ordered Griswold to assume command 
of the New Georgia Occupation Force 
at midnight of 15-16 July, and to seize 
Munda and join forces with Liversedge 
as soon as possible. Griswold took over 

38 The regimental combat team consisted of the 
161st Infantry; the 89th Field Artillery Battalion; 
A Company, 65th Engineer Battalion; and A Com- 
pany, 25th Medical Battalion. 

command as ordered. Hester reverted to 
command of the 43d Division. 

Thus by mid-July Turner and Hester, 
the two officers most responsible for the 
execution of the New Georgia tactical 
plans, had been replaced. With the offen- 
sive stalling, General Griswold was fac- 
ing his first experience in commanding 
a corps in combat. His problems were 
formidable, although some progress had 
been made. Liversedge's three battalions 
were behind schedule but had taken 
Enogai and were preparing to attack 
Bairoko. On the Munda front the 169th 
and 17 2d Infantry Regiments, also be- 
hind their schedule, had laboriously 
made their way from Zanana across the 
Barike to Laiana and the vicinity of 
Reincke Ridge and were in contact with 
the main Japanese defenses. These forces 
were obviously not adequate to break 
through and capture the airfield, but ad- 
ditional regiments were on their way. 
Aside from the difficulties presented by 
the enemy and the terrain, Griswold was 
confronted by an abnormally high rate 
of mental illness, and by the need to im- 
prove the Occupation Force supply sys- 
tem so that the regiments would be 
taken care of in the normal manner in- 
stead of by emergency air drop. Ob- 
viously, it was a case calling for general- 
ship of a high order. 


Griswold Takes Over 

General Griswold at once concluded 
that he could not mount a large-scale 
offensive against Munda until he had 
received reinforcements and reorganized 
the Occupation Force. Estimating that 
four battalions of "Munda moles well 
dug in" faced him, he planned to keep 
"pressure on slant-eye/' and to gain more 
advantageous ground for an offensive, 
by using the 43d Division in a series of 
local attacks. At the same time he would 
be getting ready for a full corps offensive 
to "crack Munda nut and allow speedy 
junction with Liversedge." 1 In the rear 
areas, Griswold and his staff set to work 
to improve the system of supply and 
medical treatment. 

The Attack on Bairoko 

Meanwhile Colonel Liversedge, after 
taking Enogai and abandoning the trail 
block, was making r eady to assault Bai- 
roko. (See Map 8.) Liversedge's opera- 

'Rad, Griswold to Harmon, 16 Jul 43, in XIV 
Corps G-g Jnl, 16 Jul 43. Unless otherwise indicated 
this chapter is based on SOPACBACOM, History 
of the New Georgia Campaign, Vol. I, OCMH; the 
orders, action rpts, jnls and jnl files of NGOF, XIV 
Corps, 43d Div, 37th Div, 35th Div, NLG, and their 
component units; USAFISPA's Daily Ir,t Summaries 
and Periodic Int Rpts; 17th Army Operations, Vol. 
II, Japanese Monogr No. 40 (OCMH); Southeast 
Area Naval Operations, Vol. II, Japanese Monogr 
No. 49 (OCMH); the published naval and air his- 
tories previously cited. 

tions against Bairoko were not closely 
co-ordinated with action on the Munda 
front. Upon assuming command Gris- 
wold directed Liversedge to submit daily 
reports, but radio communication be- 
tween Liversedge and Occupation Force 
headquarters on Rendova had been poor. 
Curiously enough Liversedge's signals 
from his Navy TBX radio could barely 
be picked up at Rendova, although the 
radio at Segi Point was able to receive 
them without much difficulty. As a result 
Liversedge had to send many messages 
through Segi Point to headquarters of 
Task Force 31 at Guadalcanal, from 
there to be relayed to Rendova, a slow 
process at best. 

In the days following the fall of Eno- 
gai, Liversedge sent patrols out to cover 
Dragons Peninsula. They made contact 
with the Japanese only once between 1 2 
and 17 July. Little information was ob- 
tained. "Ground reconnaissance," wrote 
Liversedge, ". . . was by no means all 
it should have been." Most patrols, he 
felt, were not aggressive enough, had 
not been adequately instructed by unit 
commanders, and were not properly con- 
ducted. ". . . some patrols were sent 
out in which the individual riflemen had 
no idea of where they were going and 
what they were setting out to find." 



Japanese Prisoners Captured Near Laiana Beach are escorted to division 
headquarters for interrogation. 

There was always the problem of "gold- 
bricking on the part of patrols who are 
inclined to keep their activity fairly close 
to their camp area. . . ." Patrols made 
"grave errors in distance and direction" 
and frequently were unobservant. Many 
returned from their missions unable to 
tell in what direction the streams flowed, 
whether there were fresh enemy tracks 
around a given stream, and the approxi- 
mate dimensions of swamps they had 
passed through. 2 

Prisoners might have supplied a good 
deal of information, but only two had 
been captured. Air photography, too, 

! NLG War Diary and Combat Rpt, pp. 9-10. 

might have furnished Liversedge with 
data on strongpoints, gun emplacements, 
stores, and bivouac areas, but he com- 
plained that he had received practically 
no photos. One group of obliques re- 
ceived just before the landing at Rice 
Anchorage turned out to be pictures of 
marines landing at Segi Point. Thus, ex- 
cept for the map captured on 7 July, 
Liversedge had no sound information 
on the installations at Bairoko. He was 
aware only that the Japanese were dig- 
ging in and preparing to resist. The 
Americans could only guess at Japanese 
strength at Bairoko, whither the sur- 
vivors of the Japanese garrison at Enogai 



had gone. Harmon's headquarters esti- 
mated that one Army infantry battalion 
plus two companies, some artillerymen, 
and part of the Kure 6th Special Naval 
Landing Force were defending Bairoko. 
The actual strength of the garrison is 
not clear. It consisted, however, of the 
2d Battalion, 45th Infantry, the 8th Bat- 
tery, 6th Field Artillery (both of the 6th 
Division), and elements of the Kure 6th 
Special Naval Landing Force? 

Liversedge had few more than three 
thousand men to use in the attack. The 
move of Colonel Schultz's 3d Battalion, 
148th Infantry, to Triri and the 18 July 
landing of the 4th Marine Raider Bat- 
talion at Enogai gave him a force almost 
four battalions strong, although casual- 
ties and disease had reduced the three 
battalions that made the initial landing. 
M Company and the Antitank Platoon 
of the 3d Battalion, 145th Infantry, were 
holding Rice Anchorage. The 1st and 4th 
Raider Battalions and L Company, 
145th, were at Enogai. Schultz's battalion 
and the remainder of the 3d Battalion, 
145th, were at Triri. 

Liversedge called his battalion com- 
manders together at 1500, 19 July, and 
issued oral orders for the Bairoko attack, 
which was to take place early next morn- 
ing. The Raider battalions, advancing 
some three thousand yards southwest 
from Enogai along the Enogai-Bairoko 
trail, would make the main effort. One 
platoon of B Company, 1st Raider Bat- 
talion, was to create a diversion by ad- 
vancing down the fifty-yard-wide sand- 
spit forming the west shore of Leland 

3 Japanese sources give no strength figures for 
these units. SOPACBACOM, History oE the New 
Georgia Campaign, Vol. I, Ch. V, gives approxi- 
mately two thousand, a figure which may be high. 

Lagoon. The 3d Battalion, 148th, was 
to make a separate enveloping move- 
ment. Advancing southwest from Triri 
to the trail junction southeast of Bai- 
roko, it was to swing north against the 
Japanese right flank. A and C Compa- 
nies, 1st Marine Raider Battalion, and 
elements of the 3d Battalion, 145th, 
formed the reserve at Enogai. 

Late in the day the B Company pla- 
toon took landing craft from Enogai to 
the tip of the sandspit, went ashore, and 
moved into position for the next morn- 
ing's attack. The remainder of the at- 
tacking force stayed in bivouac. From 
2000, 19 July, to 0500, 20 July, Japanese 
aircraft bombed and strafed Enogai, 
which as yet had no antiaircraft guns. 
No one was killed, but the troops had 
little rest. 

The two Raider battalions started out 
of Enogai at 0800, 20 July, and within 
thirty minutes all units had cleared the 
village and were marching down the 
trail toward Bairoko. The 1st Raider 
Battalion (less two companies) led, fol- 
lowed by the 4th Battalion and regi- 
mental Headquarters. At 0730 Schultz's 
battalion had left Triri on its enveloping 

The Northern Landing Group was at- 
tacking a fortified position. A force de- 
livering such an attack normally makes 
full use of all supporting services, arms, 
and weapons, but Liversedge's men had 
little to support them. No one seems to 
have asked for naval gunfire. Liversedge, 
who had been receiving fairly heavy air 
support in the form of bombardments 
of Bairoko, is reported to have requested 
a heavy air strike to support his assault. 
His message reached the Guadalcanal 
headquarters of Admiral Mitscher, the 



Commander, Aircraft, Solomons, too 
late on the 19th for action next day.* 

The marines definitely expected air 
support. The 4th Raider Battalion noted 
at 0900: "Heavy air strike failed to 
materialize." 5 Artillery support was pre- 
cluded by the fact that there was no 
artillery. Plindsight indicates that the six 
81-mm. mortars of the 3d Battalion, 
145th Infantry, might have been used 
in general support of the attack, but 
these weapons remained with their par- 
ent battalion. 

The Raiders advanced without meet- 
ing an enemy until 0955 when the 1st 
Battalion's point sighted four Japanese. 
When the first shot was fired at 1015, 
B and D Companies, 1st Raider Bat- 
talion, deployed and moved forward. 
Heavy firing broke out at 1045. By noon 
the battalion had penetrated the enemy 
outpost line of resistance and was in the 
outskirts of Bairoko. When D Company, 
on the left, was halted by machine gun 
fire, Liversedge began committing the 
4th Raider Battalion to the left of the 
1st. Then D Company started moving 
again. Driving slowly but steadily against 
machine gun fire, it advanced with its 
flanks in the air beyond B Company 
until by 1430 it had seized a ridge about 
three hundred yards short of the shore 
of Bairoko Harbour. Liversedge ordered 

*Maj. John N. Rentz, USMCR, Marines in the 
Central Solomons (Washington, 1952), p. m. The 
XIV Corps G— 3 Journal for 19 July contains a mes- 
sage from Liversedge, sent at 2235, 18 July, request- 
ing a twelve-plane strike on 19 July, and a "large 
strike to stand by for July 20 A M and SBD's to 
stand by for immediate call remainder of day." 
XIV Corps headquarters replied that a "large strike 
stand by" for 20 July was "impracticable." 

' 4th Mar Raider Bn Special Action Rpt, Bairoko 
Harbor, New Georgia Opn, p. 3. 

more units forward to cover D Com- 
pany's flank. These advances were made 
with rifle, grenade, and bayonet against 
Japanese pillboxes constructed of logs 
and coral, housing machine guns. The 
jungle overhead was so heavy that the 
Raiders' 6o-mm. mortars were not used. 
The platoon on the sandspit, meanwhile, 
was held up by a number of machine 
guns and was unable to reach the main- 
land to make contact with the main 

So far the marines, by attacking reso- 
lutely, had made good progress in spite 
of the absence of proper support, but 
now go-mm, mortar fire from Japanese 
positions on the opposite (west) shore 
of Bairoko Harbour began bursting 
around the battalion command posts 
and on D Company's ridge. With casual- 
ties mounting, D Company was forced 
off the ridge. By 1500 practically the 
entire force that Liversedge had led out 
of Enogai was committed and engaged 
in the fire fight, but was unable to move 
farther under the go-mm. mortar fire. 
Colonel Griffith, commanding the 1st 
Raider Battalion, regretted the absence 
of heavy mortars in the Marine battal- 
ions. Liversedge, at 1315, sent another 
urgent request for an air strike against 
the positions on the west shore of Bai- 
roko Harbour, but, as Griswold told 
him, there could be no air strikes by 
Guadalcanal-based planes on such short 
notice. With all marine units in action, 
the attack stalled, and casualties increas- 
ing, Liversedge telephoned Schultz to 
ask if his battalion could make contact 
with the marines before dark. 6 Other- 

8 All men of Headquarters Company, 4th Raider 
Battalion, were engaged in carrying Jitter cases to 
the rear. 



wise, he warned, the attack on Bairoko 
would fail. 

Schultz's battalion had marched out 
of Triri that morning in column of com- 
panies. Except for two small swamps, 
the trail was easy. By 1330 the battalion 
had traveled about 3,000 yards, passing 
some Japanese corpses and abandoned 
positions on the way, and reached the 
point where the Triri trail joined one 
of the Munda-Bairoko tracks. Here, 
about 2,500 yards south of Bairoko, the 
battalion swung north and had moved a 
short distance when the advance guard 
ran into an enemy position on high 
ground. Patrols went out to try to deter- 
mine the location and strength of the 
Japanese; by 1530 Schultz was ready to 
attack. M Company's 81 -mm. mortars 
opened fire, but the rifle companies, at- 
tempting to move against machine guns, 
were not able to advance. One officer 
and one enlisted man of K Company 
were killed; two men were wounded. 
This was the situation at 1600 when 
Schultz received Liversedge's call. 

Schultz immediately told Liversedge 
that he could not reach the main body 
before dark. A few minutes later, the 
1st Marine Raider Regiment's executive 
officer, having been dispatched to Schultz 
to tell him to push harder, arrived at the 
battalion command post. According to 
the 3d Battalion's report, the executive 
agreed that contact could not be made 
before dark and he so informed Liver- 

The group commander concluded that 
he had but one choice: to withdraw. He 
issued the order, and the marine bat- 
talions began retiring at 1700. Starting 
from the left of the line, they pulled 
back company by company. Machine 

gun and mortar fire still hit them, but 
the withdrawal was orderly. All unin- 
jured men helped carry the wounded. 
The battalion retired about five hun- 
dred yards and set up a perimeter de- 
fense on the shore of Leland Lagoon. 
When L Company of the 145th came up 
from Enogai carrying water, ammuni^ 
tion, and blood plasma, it was com- 
mitted to the perimeter. Construction of 
the defenses was impeded by darkness, 
but the task was completed and the hasty 
defenses were adequate to withstand 
some harassing Japanese that night. 

Some of the walking wounded had 
been sent to Enogai in the late after- 
noon of the 20th, and at 0615 of the 21st 
more were dispatched. Evacuation of 
litter cases began at 0830, and an hour 
later a group of Corrigan's natives came 
from Enogai to help. Carrying the 
stricken men in litters over the primitive 
trail in the heat was hard on the men 
and on the litter bearers. Liversedge 
therefore ordered that landing craft 
from Enogai come up Leland Lagoon 
and take the wounded back from a point 
about midway between Bairoko and Eno- 
gai. This evacuation was carried out, 
and by late afternoon, the withdrawal, 
which was covered by Allied air attacks 
against Bairoko, had been completed. 
All the marines were at Enogai, where 
they were joined by Schultz's battalion, 
which had retired to Triri and come to 
Enogai by boat. The Raider battalions 
lost 46 men killed, 161 wounded. They 
reported counting 33 enemy corpses, but 
estimated that the total number of en- 
emy dead was much higher. 

Once again at Enogai, the Northern 
Landing Group resumed daily patrols 
over Dragons Peninsula. 



Pressure on the Japanese 
On the Munda front, meanwhile, the 
i6gth and iy2d Regiments were engaged 
in their limited offensive to hold the 
Japanese in position and secure more 
high ground from which to launch the 
corps offensive that was to start on 25 

July. (See below, Map to.) 

The ij2d Infantry 

From 16 through 24 July the i72d 
Infantry expanded the Laiana beach- 
head. It moved west about six hundred 
yards and established a front line that 
ran for about fifteen hundred yards in- 
land from the beach near Ilangana. Dur- 
ing this period it had the support of 
tanks for the first time. Reconnaissance 
had revealed some trails in front of the 
173d that the tanks could use. Therefore 
three M3 light tanks of the 9th Marine 
Defense Battalion were assigned to each 
of the i72d's battalions, and six riflemen 
were ordered to advance with and cover 
each tank. 

In the zone of the 2d Battalion, i72d, 
on the beach, the tanks made good prog- 
ress along a jeep trail on 16 July. But 
when they reached the trail's end, their 
rate of advance slowed to about one mile 
an hour as logs, stumps, and trees caused 
constant backing, towing, and rerout- 
ing. About seventy-five yards beyond the 
2d Battalion's front lines, in an area 
where artillery fire had partly cleared 
the vegetation, the tanks sighted Jap- 
anese pillboxes. They deployed into a 
wedge formation, then fired 37-mm. high 
explosive shells. As this fire cut down 
the underbrush other pillboxes became 
visible. Japanese machine gunners man- 
ning positions in grass shacks opened 
fire, but were immediately blasted by 
canister from the tanks. Such heavy 

fire then struck the tanks that they were 
forced to close their turret hatches, but 
they found the source of much of the 
fire— a machine gun position at the base 
of a banyan tree. The marines shot at 
this position for some time, but as they 
killed one gunner, his replacement 
would bound forward from the rear, 
man the gun, and keep shooting until 
he was killed. At length the tanks de- 
stroyed the gun, drove the surviving 
crew members into a nearby pillbox, 
pulled up close, and demolished three 
pillboxes with short-range fire. Troops 
of the 2d Battalion then moved forward 
to grenade the wreckage. 

The three tanks operating with the 
3d Battalion, to the right of the 2d, had 
less success, as the ridges in that zone 
were so steep that the tanks could not 
elevate or depress their guns enough to 
hit the enemy positions. 

The destruction of the pillboxes near 
the shore gave the troops an opportunity 
to inspect the type of defenses they would 
have to overcome before they could 
take Munda. The pillboxes were not 
concrete, as had been feared, but were 
made of coconut logs and coral. From 
ten to twelve feet square, they had three 
or four layers of logs banked with six 
to eight feet of weathered coral. About 
ten feet from floor to ceiling, they were 
dug into the earth so that only two or 
three feet of pillbox projected above 
the ground. Each had several firing slits 
for riflemen as well as a firing platform 
for a heavy machine gun. Outside were 
foxholes among banyan and mahogany 
trees. Trenches connected all positions, 
which were well camouflaged. Besides 
employing terrain contours for conceal- 
ment, the Japanese used earth, grass, 



Pillbox Made or Coconut Logs and Coral near Munda Airfield 

vines, palm fronds, and leaves to such 
good effect that the American soldiers 
might receive fire from a pillbox and 
still not be able to see it. Soldiers of the 
43d Division remarked that the Japanese 
positions were easier to smell than see. 
As usual, the Americans reported the 
presence of many snipers in trees, but 
these reports had little basis in fact. No 
one ever seems to have actually seen one. 

The tanks attacked again on the 17 th, 
but lack of tank-infantry co-ordination 
hampered their efforts. The Marine tanks 
and the Army infantry had not trained 
together. Foot soldiers had no sure means 
of communicating with the tanks when 
they were closed up for action. Tank 
crews, with hatches closed, could see 

very little in the jungle. The tankers ut- 
tered the classic complaint that the rifle- 
men did not give them proper support 
and protection, while the infantrymen 
claimed that the tanks did not always 
press forward to support them. Doubt- 
less both accusations were based on 

Japanese antitank tactics, practically 
nonexistent at first, improved each day, 
for staff officers had hurried down 
from Rabaul to instruct Sasaki's men in 
methods of dealing with tanks. The 
Japanese used mines, flame throwers, 
Molotov cocktails, and fuzed charges 
of TNT against the tanks, but appar- 
ently had no antitank guns. After two 
tanks were permanently disabled on 17- 



18 July, General Griswold withdrew 
the other tanks from the front to permit 
repairs. He ordered the gth Marine De- 
fense Battalion tank commander to re- 
connoiter for terrain suitable for tank 
action, and at the same time requested 
that the Tank Platoon of the 10th Ma- 
rine Defense Battalion, then in the Rus- 
sells, be sent to New Georgia. 

Kelley Hill 

In the 169th Infantry's zone farther 
north, the 3d Battalion's seizure of 
Reincke Ridge was being exploited. The 
2d Battalion was able to capture the hill 
immediately north of Reincke Ridge, 
and on 15 July Maj. Joseph E. Zimmer, 
commanding the 1st Battalion, recon- 
noitered the high ground (Kelley Hill) 
four hundred yards southwest of Rein- 
cke Ridge in preparation for an attack. 

At 0830 the next day, 16 July, the 
155-mm. howitzers of the 136th Field 
Artillery Battalion and the 3d Battalion's 
mortars put fire on the objective. At 
the same time the 1st Battalion, fortified 
by hot coffee and doughnuts, passed 
through the 3d Battalion's lines and ad- 
vanced to the attack. One platoon from 
C Company, carrying .30-caliber light 
machine guns, struck out down the west 
slope of Reincke Ridge and up the east 
slope of Kelley Hill, seized the crest, and 
set up machine guns to cover the ad- 
vance of the battalion's main body, 
which was to envelop Kelley Hill from 
the south. The whole effort was blood- 
less. The battalion's advance elements 
climbed the hill without meeting any 
opposition. They found only empty 
pillboxes and abandoned foxholes. By 
1530 the entire battalion was on the 
ridge top. The men found they could 

look west and see the waters south of 
Munda Point, although the airfield was 
hidden from view. Because natives had 
formerly dug yam gardens on the ridge, 
there was an open area about 75 by 150 
yards. Zimmer's men, using Japanese 
positions when possible, started building 
an all-round defense in the clearing. 
Automatic rifles, machine guns, and 
Mi 903 and Mi rifles were posted on the 
line, with mortars in supporting posi- 
tions in rear. 

There was a brush with a Japanese 
patrol at 1650, and before dark, when 
the emplacements were still incomplete, 
Japanese artillery and mortar fire struck 
the battalion. Fourteen men died, in- 
cluding 1st Lt. John R. Kelley, in whose 
memory the hill was named. Just fifteen 
minutes after midnight part of the 3d 
Battalion, zzgth Infantry, now com- 
manded by Captain Kojima, assaulted 
the hill from positions on Horseshoe 
Hill. Beaten off, Kojima tried twice 
more against the right (north) and rear 
(east) but failed to dislodge Zimmer's 

The 1st Battalion held to the ridge, 
but as day broke on 17 July the troops 
realized that their situation was not en- 
viable. That the Japanese were still 
active was indicated by their resistance 
to an attempt by the 2d Battalion to 
drive into the draw between Reincke 
Ridge and Kelley Hill. This attempt 
was beaten back. The 1st Battalion's 
rations and ammunition were running 
low; the battalion surgeon had no medi- 
cal supplies. And when Japanese ma- 
chine guns fired on a party carrying 
twenty wounded men to the rear and 
forced it to return west to Kelley Hill, 
the men of the battalion knew that they 



were virtually isolated. Fortunately the 
telephone line to the regimental com- 
mand post was still operating, and Major 
Zimmer was able to keep Colonel Hol- 
land informed on his situation. As the 
hot day wore on, the supply of water 
dwindled. Some men left their positions 
to drink from puddles in shell holes. 
Eight of those who thus exposed them- 
selves were wounded by Japanese rifle- 
men. In midafternoon succor came. A 
party of South Pacific Scouts, accom- 
panied by Capt. Dudley H. Burr, the 
regimental chaplain, escorted a supply 
party through to Kelley Hill. The party 
brought ammunition, rations, water, 
blood plasma, litters, and orders from 
Holland to hold the hill. The wounded 
were carried out. The unwounded on 
Kelley Hill, securely dug in, made ready 
to meet the Japanese night attack which 
they had reason to expect. 

The Enemy Counterattacks 

Up to now, Japanese ground troops 
had harried the Americans at night with 
local attacks, but had not attempted 
any large co-ordinated offensives. They 
had manned their defensive positions, 
fired at the American infantry, and had 
received bombs, shells, and infantry as- 
saults without retaliating very actively. 
This quiescence, so different from en- 
emy reactions during the Guadalcanal 
Campaign, puzzled the American com- 
manders. General Sasaki was well aware 
that only offensive action would destroy 
the Allied forces on New Georgia, and 
he had brought the 13th Infantry to 
Munda from Kolombangara for that 

Sasaki ordered the 13th, acting in con- 
cert with as much of the 229th Infantry 
as he could spare from the defenses east 

of Munda, to assemble on the upper 
reaches of the Barike, fall upon the Al- 
lied flank and rear, and destroy the 
whole force. 7 The 13th Infantry, having 
completed its march from Bairoko, as- 
sembled on the upper Barike on 15 July. 
It claims to have attacked the 43d Divi- 
sion's right flank on that date, a claim 
that is not supported by the 43d Division 
records. Two days later the 13th made 
ready to attack from the upper Barike. 

In the afternoon of the 1 7th American 
patrols operating on the practically open 
right flank reported that an enemy col- 
umn, 250-300 men strong, was moving 
eastward. A platoon from the 43d Cav- 
alry Reconnaissance Troop went out to 
ambush the column but failed to inter- 
cept it. It was obvious that the Japanese 
had some sort of offensive action in 

It was equally obvious that the Allied 
forces in front of Munda were in a vul- 
nerable position. Their right flank was 
in the air; the front line positions were 
exposed to envelopment from the north. 
The Japanese reinforcement route from 
Bairoko was still open, and 43d Division 
-rear installations, strung out from Za- 
nana to the front, were unguarded ex- 
cept for local security detachments. 
Movement was slow along the Munda 
Trail; the track north from Laiana was 
not yet completed. It would thus be 
difficult to send speedy reinforcement to 
any beleaguered unit. A resolute, skill- 
ful attack by the 13th Infantry, such as 
Sasaki had planned, could destroy the 
43d Division's rear installations, cut 
the line of communications from Zanana 
to the front, and if co-ordinated with the 

7 This order probably accounts tor the with- 
drawal of part of the 229th from Kelley Hill. 



efforts of the 229th Infantry might sur- 
round the American regiments on the 
front lines. 

Captain Kojima was ready to do his 
part. He had prepared another attack 
against Kelley Hill. At 0015, 18 July, 
Japanese machine guns north of Kelley 
opened fire. They covered the advance 
of riflemen who were attempting an 
assault against the west slope of Kelley 
Hill. The 1st Battalion fired at the Jap- 
anese infantry with all weapons that 
would bear, including two captured Jap- 
anese machine guns. Tracers from Ko- 
jima's machine guns revealed their loca- 
tion, and 3d and 1st Battalion mortar 
crews put their fire on the Japanese 
positions to the north. Kojima's first at- 
tack failed. His men pulled back, re- 
grouped, and tried again, this time from 
the north. They succeeded in seriously 
threatening the line. The broken ground 
on the north slope of Kelley Hill pro- 
vided some cover from the fire of one 
of the machine guns that was supposed 
to sweep the area. The Japanese, taking 
advantage of the dead space, crawled 
within grenade-throwing range of the 
northern line of the 1st Battalion. But 
mortar fire killed some of them and 
forced the others to withdraw. The 1st 
Battalion reported counting 102 Jap- 
anese bodies on the slopes of Kelley Hill 
after daybreak. 

A predawn attack by the 2d Battalion, 
229th Infantry, against the beach posi- 
tions of the 3d Battalion, 103d Infantry, 
in the i72d's sector, was readily re- 
pulsed. 8 

Elsewhere on the night of 17-18 

8 The 3d Battalion, 103d, was attached to the 
173d Infantry. 

July the Japanese caused alarms and up- 
roar. They launched simultaneous raids 
against the engineer and medical biv- 
ouacs and the 43d Division command 
post at Zanana. Near one of the Barike 
bridges they ambushed a party taking 
wounded of the 169th to the rear, then 
attacked the hasty perimeter set up by 
the party and killed several of the 

The attacks against the engineer and 
medical bivouacs were easily beaten off, 
but at the command post the raiders' 
first onslaught carried them through the 
security detachment's perimeter and into 
the communication center where they 
ripped up telephone wires and damaged 
the switchboard before being chased off. 
The division artillery liaison officer, 
Capt. James Ruhlen, called for support- 
ing fire from the 136th Field Artillery 
Battalion. Adjusting by sound, he put 
fire on a nearby hill where the Japanese 
were thought to be emplacing mortars 
and laid a tight box barrage around the 
command post. This fire was continued 
throughout the night. During the action 
Lt. Col. Elmer S. Watson, 43d Division 
G-3, was wounded. Maj. Sidney P. Mar- 
land, Jr., his assistant, took his place. 

Shortly after receiving word of the at- 
tack, General Griswold ordered a bat- 
tery of artillerymen from Kokorana to 
Zanana to protect the command post, 
and on his orders Colonel Baxter se- 
lected the 1st Battalion of his 148th In- 
fantry to move from Rendova to Za- 
nana at daybreak. 

The rjth Infantry then withdrew to 
the north. It had caused a few casualties 
but accomplished very little, certainly 
not enough to justify its trip from Kol- 
ombangara. As might be expected, Gen- 



eral Sasaki was disappointed. 9 Reincke 
Ridge, Kelley Hill, and Laiana beach- 
head remained in American hands. 

Preparations for the Corps Offensive 

Commitment of the 37th Division 

General Griswold, preparing for his 
corps offensive, needed fresh troops at 
the front. On 18 July he ordered Colo- 
nel Baxter to advance west with the 2d 
Battalion of his 148th Infantry and re- 
lieve the 169th Infantry as soon as pos- 
sible. Baxter, whose 1st and 2d Battal- 
ions had arrived at Zanana that morning, 
effected the relief by 21 July after being 
delayed by Japanese detachments at the 

After the i6gth's relief, regimental 
command changed again. Colonel Hol- 
land took over his old regiment, the 
145th, while Lt. Col. Bernard J. Lind- 
aner succeeded to command of the 169th. 
Lindauer's regiment returned to Ren- 
dova for rest and reorganization. Its 3d 
Battalion, after receiving 212 replace- 
ments, was sent into reserve at Laiana 
on 24 July. 

By 23 July the major part of the 37th 
Division had arrived at New Georgia 
and was either in action or ready to be 
committed. Present were Division and 
Division Artillery Headquarters; the 
145th and 148th Infantry Regiments less 
their 3d Battalions, which were under 

'Japanese records do not indicate just what the 
main body of the ijth actually did during the 
period j 7—1 g July- The various raids could not 
have been the work of the entire unit. The main 
body apparently never got into action at all. The 
170 hungry survivors of Major Hara's Vim garrison 
may have caused some of the trouble to the Ameri- 
cans, for on 18 or 19 July they reached Munda after 
marching overland from Viru and infiltrating the 
American lines from the rear. 

Liversedge; the 135th and 136th Field 
Artillery Battalions; the 37th Cavalry 
Reconnaissance Troop; and the signal, 
quartermaster, ordnance, engineer, and 
medical units (except B Company, 
117th Engineer Battalion, and B Com- 
pany, 11 2th Medical Battalion). 

General Griswold, on 22 July, directed 
General Beightler to resume command 
at noon of all his units then on New 
Georgia except the 136th Field Artillery 
Battalion. To the 37th Griswold at- 
tached the 161st Regimental Combat 
Team less its artillery, and the 169th 
and i92d Field Artillery Battalions of 
the 43d Division. The 136th Field Artil- 
lery Battalion was serving as part of 
corps artillery. The other three organic 
and attached artillery battalions were 
under the 37th Division for direct fire 
support missions only; for all others they 
would be controlled by corps artillery, 
now commanded by General Barker. 

Griswold, reshuffling units for the 
offensive, set the boundary between divi- 
sions along an east-by-south-west-by- 
north line approximately thirteen hun- 
dred yards north of Ilangana. The 43d 
Division was on the left (south), with 
the 103d and i72d Regiments in line 
from south to north. 10 The 17 2d moved 
right to establish contact with the 37th 
Division's left. 

The 37th Division, assigned an indefi- 
nite frontage north of the 43d Division, 
gave the 145th Infantry a narrow front 
of 300 yards on the left, because only the 
2d Battalion, 145th, which had been 
covering the gap north of the i72d In- 
fantry, was immediately available. The 

"The 2d Battalion, 103d, having been relieved 
by the 1st Battalion, had come up from Wickham 

Soldiers of the i6ist Infantry debarking from an LCI, New Georgia, 22 July 1943- 



ist Battalion was still holding the high 
ground taken over from the 169th In- 
fantry. The 161st Infantry was given a 
500-yard front in the center. One of its 
battalions constituted the corps reserve. 
The 148th Infantry was put on the right, 
with no definite frontage, and assigned 
the responsibility for protecting the 
corps' right flank and rear. 

All units had moved into position by 
24 July. The 161st Infantry, whose trans- 
fer had been approved by General Har- 
mon, had arrived at Baraulu Island on 
21 July, moved to New Georgia the next 
day, and suffered its first casualties of 
the campaign when two captains of the 
regimental staff were killed on recon- 
naissance. On 23 July the regiment 
moved to assembly areas in preparation 
for the offensive. Most of the 1 Gist's zone 
of action lay north of the high ground 
taken by the 169th Infantry. 

The corps line of departure ran north- 
west from a point near Ilangana. In the 
i6ist's zone, it lay about three hundred 
yards west of the assembly areas, and 
ran over Horseshoe Hill. Colonel Dal- 
ton, who had taken over command of the 
regiment in the closing days of the 
Guadalcanal Campaign, sent out patrols 
to reconnoiter for the line of departure. 
These patrols were stopped short of the 
line by Japanese on a ridge that formed 
part of the northeast slope of Horseshoe 
Hill, and returned to report to Dalton 
that there were two pillboxes on the 
ridge. A reinforced platoon went out to 
deal with the enemy. This platoon came 
back and claimed the destruction of two 
positions but reported the presence of 
several more. Because Beightler did not 
want to commit the regiment to general 
action before 25 July, he ordered Dalton 

to use one rifle company to clear the 
ridge on 24 July. I Company, supported 
by M Company's 81 -mm. mortars, at- 
tacked and reported knocking out two 
more pillboxes, apparently by killing 
the occupants. But I Company also re- 
ported the presence of a dozen more pill- 
boxes. Before nightfall, patrols reported 
that the Japanese had reoccupied the 
two positions I Company had attacked. 
Thus just before D Day the 161st Infan- 
try was aware that a strong enemy posi- 
tion lay camouflaged between it and the 
line of departure. 


In the days following his assumption 
of command, General Griswold and his 
staff were deeply occupied with admin- 
istrative as well as tactical matters. Re- 
inforcements from the 25th and 37th 
Divisions had to be received and as- 
signed. The supply system was over- 
hauled; medical services were improved. 

General Griswold immediately desig- 
nated Barabuni Island as supply dump 
for the 43d Division, Kokorana for the 
37th. Ships from Guadalcanal would 
land equipment and supplies in these 
islands, whence landing craft would 
transport them through the barrier is- 
lands to Laiana or to other positions on 
the barrier islands. 

Hester's move to Laiana was pay- 
ing dividends. Although low, swampy 
ground had at first slowed construction 
of the trail from Laiana north to the 
Munda Trail, six hundred yards had 
been built by 17 July, and on 20 July the 
whole trail was opened to motor traffic. 
As a result, Hester reported, his regi- 
ments would no longer need to be sup- 
plied from the air. The 43d Division 



command post moved from Zanana to 
Laiana on 21 July. At the same time 
most of the 43d Division's service instal- 
lations moved to Laiana. Two-lane roads 
were built within the dump areas, and 
additional trails out of Laiana, plus more 
trails to the various regiments, were also 
built. Bulldozer operators working in- 
land received fire from enemy rifle- 
men on occasion. After one driver was 
wounded, the engineers fashioned shields 
for the bulldozers with steel salvaged 
from wrecked enemy landing craft. A 
D-4 and a much heavier D-7 bulldozer 
that came in with A Company, 65th En- 
gineer Battalion, on 23 July, speeded 
construction of a trail to the 161st In- 
fantry and of lateral trails in the 37th 
Division's area. With the roads built it 
was possible to assemble supplies close 
behind the infantry regiments and to 
plan their systematic delivery in the 

The XIV Corps and its assigned units 
also undertook the improvement of med- 
ical care. Several hours after he assumed 
command Griswold asked Harmon to 
send the 250-bed 17th Field Hospital 
from Guadalcanal to Rendova at once. 
Harmon approved. Because of physical 
frailty some medical officers had become 
casualties themselves, and the resulting 
shortage prevented careful supervision 
and handling of casualties. Griswold 
asked Harmon for fifteen medical officers 
physically able to stand the rigors of field 
service. To make sure that casualties 
being evacuated from New Georgia re- 
ceived proper medical attention during 
the trip to Guadalcanal, the corps sur- 
geon arranged with naval authorities 
for a naval medical officer to travel on 
each LST carrying patients. 

Finally, all units benefited by the 43d 
Division's experience in dealing with 
war neurosis. Rest camps, providing hot 
food, baths, clean clothes, and cots, were 
established on the barrier islands, and 
Colonel Hallam tried to see to it that 
more accurate diagnoses were made so 
that men suffering from combat fatigue 
were separated from true neurotics and 
sent to the camps. 

Air Support 

Air support of the New Georgia oper- 
ation had been generally good, and the 
scale of bombing was increasing. Com- 
pletion of the Segi Point field on 10 
July and full employment of the Rus- 
sells fields made it possible for fighters 
to escort all bombing missions. These 
missions could therefore be executed in 
daylight with resulting increases in ac- 
curacy. South Pacific air units were able 
to put more planes in the air at one time 
than ever before. Regular strikes against 
the Shortlands and southern Bougain- 
ville were intensified. 

Allied lighters providing the 0700 to 
1630 cover for the New Georgia Occupa- 
pation Force also escorted the almost 
daily bombing attacks against Munda, 
Bairoko, and Vila. Fighter operations 
were proving especially effective in pro- 
tecting the beachheads and shipping. 
On 15 July some seventy-five Japanese 
bombers and fighters were intercepted 
by thirty-one Allied fighters, who re- 
ported knocking down forty-five enemy 
craft at a cost of three American planes. 
Thereafter Japanese aircraft virtually 
abandoned daylight attacks against Ren- 
dova and New Georgia and confined 
their efforts to nocturnal harassment. 

Bombing and strafing missions in sup- 



Bombing of Munda Airfield, Early Morning, 12 July 1943. Photograph taken 
from Kokorana Island. 

port of the ground troops were numerous 
and heavy, considering the number of 
aircraft in the South Pacific. On 16 July 
37 torpedo bombers and an equal num- 
ber of dive bombers struck at Lambeti 
with thirty-six 1,000-pound, eighteen 
2,000-pound, and eighty-eight 500-pound 
bombs at 0905. The strike was followed 
by another against Munda by 36 SBD's 
and TBF's. These dropped twelve 1,000- 
pound and twelve 2,000-pound bombs 
at 1555. On 19 July 20 TBF's and 18 
SBD's hit at positions north of Munda, 
and the next day 36 SBD's dropped 
1,000-pound bombs at suspected gun 
positions north of Lambeti. Two days 
later 36 SBD's and 18 TBF's again 
bombed the Munda gun positions, which 
were struck once more by 16 SBD's on 
23 July. On 24 July, the day before the 

corps offensive began, 37 TBF's and 36 
SBD's with a screen of 48 fighters 
dropped thirty-seven 2,000-pound and 
thirty-six 1,000-pound bombs on Bai- 
roko in the morning. In late afternoon 
18 SBD's and 16 TBF's hit Munda and 
Bibilo Hill. 

Most of the aircraft flying these mis- 
sions were piloted by marines. It will be 
noted that this air support was, accord- 
ing to then current Army doctrine, di- 
rect air support. Most of these missions 
were flown as part of "a combined effort 
of the air and ground forces, in the bat- 
tle area, to gain objectives on the imme- 
diate front of the ground forces." 11 But 

11 See FM 100-20, Command and Employment of 
Air Power (1943), p. 16. See also TM 80-205, 
Dictionary o£ United States Army Terms (1944), p. 



as most of the targets were several thou- 
sand yards from the front lines, this 
was not close air support, which was 
defined after the war as "attack by air- 
craft on hostile ground or naval targets 
which are so close to friendly forces as 
to require detailed integration of each 
air mission with the fire and movement 
of these forces." 12 South Pacific com- 
manders, including General Harmon, 
had hoped to make extensive use of close 
air support on New Georgia, and a few 
close air support missions such as that 
requested by Colonel Holland had been 
executed, but they were difficult for the 
air forces to execute and dangerous to 
the ground troops. There was, at that 
time, no systematized target marking 
system nor any good means of radio com- 
munication between the front lines and 
the aircraft. The Thirteenth Air Force 
had no tactical air communications 
squadron. The dense jungle and rolling 
terrain where the troops were operating 
had so few landmarks that pilots could 
not easily orient themselves. Nor could 
the ground troops orient themselves any 
more easily. Panels marking the front 
lines could scarcely be seen from the air. 
Enemy positions could rarely be identi- 
fied by spotters in observation planes or 
by air liaison parties on the ground. Be- 

13 SR 320-5—1, Dictionary o£ United States Army 
Terms, (Aug 50), p. 48. 

cause maps were inexact, and the troops 
had difficulty in locating themselves pre- 
cisely, bombing missions executed close 
to the front lines resulted in casualties 
to American troops. Three soldiers of 
the i72d Infantry were killed in that 
way on 16 July. For these reasons close 
air support was seldom used in any of 
the Cartwheel operations. The direct 
air support on New Georgia was, how- 
ever, of great value, and General Gris- 
wold had every intention of following 
Harmon's order that the New Georgia 
operation employ air support to the 
maximum degree. 

Griswold, in the nine days following 
his assumption of command, had im- 
proved supply and evacuation on New 
Georgia. In spite of the failure at Bai- 
roko, the tactical position, too, had im- 
proved. The tired 169th had been re- 
lieved, and fresh 25th and 37th Division 
regiments were ready to enter the fight. 
The troops had repulsed a counterat- 
tack, improved their position by seizing 
high ground, and now held a southeast- 
northwest line about three thousand 
yards from the east end of Munda field. 
The XIV Corps could look forward 
to receiving the same resolute, effective 
air and naval support that had aided the 
43d Division. With the logistic and tac- 
tical situations of his troops thus im- 
proved, and sure of ample air and naval 
support, Griswold was ready to attack 


XIV Corps Offensive 

The plan for the XIV Corps' drive 
against Munda was completed shortly 
after Griswold took over. 1 Col. Eugene 
W. Ridings, Griswold's assistant chief of 
staff, G-g, flew to Koli Point to confer 
with General Harmon and Admiral Wil- 
kinson on naval gunfire and air support. 
Ridings also asked for, and obtained, a 
better radio (SCR 193) for Liversedge, 
to improve communications between 
him and Griswold. Harmon stressed the 
importance of submitting a precise plan 
for air support to Admiral Mitscher. 
Dive bombers would naturally be the 
best for close work, while mediums and 
heavies should be used for area bombing, 
he asserted. Harmon agreed to send in 
more tanks at Griswold's request. 2 


The American Plan 

Naval support plans called for a seven- 
destroyer bombardment of Lambeti 
Plantation shortly before the infantry's 
advance. Comdr. Arleigh A. Burke, the 
destroyer division commander, came to 
Rendova on 23 July to view Roviana La- 

1 Unless otherwise indicated this chapter is based 
on the same sources as Chapter VIII. 

1 It was at this conference that he approved the 
transfer of the 161st Regimental Combat Team to 
New Georgia. 

goon and select visual check points. Air 
support for the offensive would include, 
besides the normal fighter cover, pattern 
bombing by multiengine planes in front 
of the 43d Division about halfway be- 
tween Ilangana and Lambeti Plantation. 
Single-engine planes would strike at po- 
sitions north and northeast of Munda 
field. Artillery spotting planes and liai- 
son planes would be on station con- 

Artillery support would be provided 
by Barker's artillery from its island posi- 
tions. Plans called for fairly standard 
employment of the field artillery, pro- 
viding for direct and general support of 
the attack, massing of fires in each in- 
fantry's zone of advance, counterbattery 
fire, and the defense of Rendova against 
seaborne and air attack. One 105-mm. 
howitzer battalion was assigned to direct 
support of each regiment, one 155-mm. 
howitzer battalion to general support of 
each division. Except for specific direct 
and general support missions, all artil- 
lery would operate as the corps artillery 
under Barker. The XIV Corps had 
neither organic artillery nor an artillery 
commander. 3 

Griswold's field order, issued on 22 

1 See Miller, Guadalcanal: The First Offensive, 
pp. 218-19. 



July, directed his corps to attack vigor- 
ously to seize Munda airfield and Bibilo 
Hill from its present positions which 
ran from Ilangana northwest for about 
three thousand yards. (Map 10) The 37th 
Division was to make the corps' main 
effort. Beightler's division was to attack 
to its front, envelop the enemy's left 
(north) flank, seize Bibilo Hill, and 
drive the enemy into the sea. At the 
same time it would protect the corps' 
right flank and rear. The 43d Division 
was ordered to make its main effort on 
the right. Its objectives were Lambed 
Plantation and the airfield. Liversedge's 
force, depleted by the abortive attack 
on Bairoko, was to continue patrolling 
and give timely information regarding 
any Japanese move to send overland re- 
inforcements to Munda. The gth De- 
fense Battalion's Tank Platoon would 
assemble at Laiana under corps control. 
The 1st and 2d Battalions, i6gth In- 
fantry, at Rendova, constituted the corps 
reserve. 4 

All units were ordered to exert un- 
ceasing pressure on the enemy. Isolated 
points of resistance were not to be al- 
lowed to halt the advance, but were to 
be bypassed, contained, and reduced 
later. Griswold ordered maximum use of 
infantry heavy weapons to supplement 

4 Griswold issued his attack order as NGOF FO 1, 
22 July 43» although Hester had issued NGOF FO 
1 in June. 

The 37th Division was less the 129th Regimental 
Combat Team and the 3d Battalions, 145th and 
148th Infantry Regiments, and was reinforced by 
the 161st Regimental Combat Team (less its artil- 
lery) and a detachment of South Pacific Scouts. 

The 43d Division was less nearly all its head- 
quarters, whose officers were filling most of the posts 
in Occupation Force headquarters, and less two bat- 
talions of the 169th Infantry and the 1st Battalion, 
103d Infantry. 

artillery. Roads would be pushed for- 
ward with all possible speed. 

D Day was set for 25 July. The thirty- 
minute naval bombardment was to start 
at 0610, the air bombing at 0635. The 
line of departure, running northwest 
from Ilangana, was practically identical 
with the American front lines except in 
the zone of the 161st Infantry where the 
existence of the Japanese strongpoint 
east of the line had been determined 
on 24 July. 

The XIV Corps was thus attempting 
a frontal assault on a two-division front, 
with the hope of effecting an evelop- 
ment on the north. In the initial attack 
it would employ two three-battalion regi- 
ments (the 161st and the 17 2d) and three 
two-battalion regiments (the 103d, the 
145th, and the 148th). 

Enemy Positions and Plans 

On 22 July the Japanese front line 
ran inland in a northwesterly direction 
for some 3,200 yards. This line was 
manned by the entire 229th Infantry, 
and at the month's end the 2d Battalion, 
■230th Infantry j was also assigned to it. 
In support were various mountain artil- 
lery, antitank, antiaircraft, and auto- 
matic weapons units. 5 The positions 
were the same complex of camouflaged 
and mutually supporting pillboxes, 
trenches, and foxholes that had halted 
Hester in midmonth. The pillboxes 
started near the beach at Ilangana and 
ran over the hills in front of the 
103d, i72d, 145th, and 161st Regiments. 

5 These included the Antitank Battalion, j8th 
Division; 2d Independent Antitank Battalion; a de- 
tachment of the 2d Battalion, 90th Independent 
Mountain Artillery Regiment; thirteen 7.7-mm. 
machine guns, and two 75-mm. antiaircraft units. 


22 July- 4 Auflusl 1943 

U.S. position, date indicateo 

Terrain features and military positions topproi) 
Form tint interval SO fast 

50O S00 1000 YARDS 

MAP 10 

F. Ttmple 



A particularly strong series lay on a 
tangled set of jungled hills: Shimizu 
Hill in front of the i72d Infantry, and 
Horseshoe Hill (so named from its con- 
figuration) in front of the 145th and 
161st Regiments. Horseshoe Hill lay 
northwest of Kelley Hill and west of 
Reincke Ridge. East of Horseshoe Hill 
lay the Japanese pocket discovered by the 
161st. The pocket lay on a north-south 
ridge that was joined to Horseshoe Hill 
by a rough saddle. The pillbox line ter- 
minated at about the northern boundary 
of the 161st Infantry. When the 2d Bat- 
talion, 230th Infantry, was committed it 
did not occupy carefully prepared posi- 
tions. From the end of the pillboxes the 
line ran west to the beach, and this north 
flank does not seem to have been strongly 

XIV Corps headquarters still esti- 
mated that four enemy battalions faced 
it; three at Munda and one at Bairoko. 
This was a fairly accurate estimate of 
strength on the enemy line, but Sasaki 
had an ace up his sleeve— the 13 th In- 
fantry. This regiment, which was not in 
full strength, was stationed on the Amer- 
ican right flank about 4,900 yards west 
by north from Ilangana. Sasaki's plans 
to use his ace were similar to his earlier 
plans. On the same day that Griswold 
issued his field order, Sasaki directed 
Colonel Tomonari to attack the Ameri- 
can right flank in the vicinity of Horse- 
shoe Hill on 23 July, then drive east 
along the Munda Trail. But the Ameri- 
cans struck before Tomonari made his 

Ilangana and Shimizu Hill: 
The 43d Division 
In the 43d Division's zone, the offen- 
sive began as scheduled on the morning 

of 25 July. For once the weather was 
favorable. D Day dawned fair and clear, 
with visibility as good as could be ex- 
pected in the jungle. 

Naval gunfire, air, and artillery prepa- 
rations went off as scheduled. Com- 
mander Burke's seven destroyers had 
sailed up from Tulagi. At 0609 the two 
screening destroyers fired the first of 
four thousand 5-inch shells at Lambeti 
Plantation; these were followed by the 
main group at 0614. Visibility to sea- 
ward was good, but the morning haze 
still hung over Lambeti Plantation. Fif- 
teen minutes later visibility had im- 
proved but now the target area was 
partly obscured by smoke and dust raised 
by the bombardment. 6 Firing ceased at 

From 0630 to 0700, 254 aircraft un- 
loaded 500,800 pounds of fragmentation 
and high explosive bombs on their tar- 
get area, a i,5oo-by-25o-yard strip begin- 
ning about 500 yards west of the 103d 
Infantry's front lines. No corps artillery 
concentrations were fired on 25 July, 
but the 43d Division's supporting artil- 
lery began before 0*700 the first of more 
than 100 preparations that were fired 

8 When the infantrymen later reached Lambeti 
Plantation they found that although the bombard- 
ment had done extensive damage many positions, 
which could have been destroyed only by direct 
hits, remained intact. The theoretical density of 
this shelling was 70 rounds per 100 square yards. 
Admiral Wilkinson, who thought the target area 
was too far west of the front lines, later observed 
that 200 rounds per 100 square yards would be re- 
quired to achieve complete destruction. Because of 
the difficulty in distinguishing targets in morning 
haze, Commander Burke recommended that shore 
bombardments should not start earlier than twenty 
minutes before sunrise. See CTF 31 (Com III 
AMPHFOR) Action Rpt for Morning 25 Jul 43, 
The Bombardment of Munda, 3 Sep 43, in Off of 
Naval Reds and Library; and ONI USN, Operations 
in the New Georgia Area, pp. 53—54. 



that day. The 103d and ir^d Field Artil- 
lery Battalions fired more than 2,150 
105-mm. howitzer shells; the 155-mm. 
howitzers of the ig6th Field Artillery 
Battalion threw 1,182 rounds at the 

With the din subsiding as the artillery 
shifted its fire to positions farther west, 
the infantrymen of the 43d Division 
moved to the attack at 0700. In the 17 2d 
Infantry's zone the 2d and 3d Battalions 
on the left and right attacked westward 
against Shimizu Hill. But by 1000 they 
had run into the enemy pillbox line 
and halted. Colonel Ross then requested 
tanks, got some from the corps reserve, 
and attacked again. By 1430 three tanks 
were disabled, and the attack stalled. A 
little ground had been gained on the 
regimental left. 

The 103d Infantry, now commanded 
by Colonel Brown, attacked alongside 
the i72d with little more success. 7 The 
3d Battalion, on the left, pushed for- 
ward against machine gun and mortar 
fire, but immediately hit the Japanese 
line and stopped. The battalion at- 
tempted to move around the pillboxes 
but found that this maneuver took its 
men into other machine gun fire lanes. 

The 2d Battalion, 103d, in the center 
of the 43d's zone, did better. It moved 
forward two or three hundred yards 
against light opposition. By 1040 E Com- 
pany's leading elements had advanced 
five hundred yards. The company kept 
moving until noon, when it had reached 
the beach near Terere. Here it set up a 
hasty defense position. But the com- 

1 Colonel Brown, formerly commander o£ Che ad 
Battalion, took over regimental command when 
Colonel Hundley replaced the 43d Division chief of 
staff on 22 July. 

panies on either flank had not been able 
to keep up, and the Japanese moved in 
behind E Company to cut the telephone 
line to battalion headquarters. 

To exploit E Company's break- 
through, General Hester took the 3d Bat- 
talion, 169th Infantry, out of division 
reserve and ordered it to push through 
the same hole E Company had found. 
But the Japanese had obviously become 
aware of the gap, and as the 3d Battalion 
marched to the line of departure it was 
enfiladed by fire from the south part of 
Shimizu Hill and from the pillboxes to 
the south. It halted. Five Marine tanks 
were then ordered to push over Shimizu 
Hill but could not get up the steep 
slopes. When three of them developed 
vapor lock all were pulled back to Lai- 
ana. In late afternoon the E Company 
commander decided to abandon his ex- 
posed, solitary position, and E Company 
came safely back through the Japanese 
line to the 2d Battalion. 

North of the 43d Division the 37th 
Division had made scant progress. 8 Thus 
the first day of General Griswold's of- 
fensive found the XIV Corps held for 
little gain except in the center of the 43d 
Division's line. 

The 43d Division was weakened by al- 
most a month's combat, and its reduced 
strength was spread over a long, irregu- 
lar, slanting front. It was obvious that 
combat efficiency would be increased 
by narrowing the front, and this could 
be done by advancing the left and 
straightening the line. Consequently 
Hester's plan for 26 July called for the 
i72d to stay in place while the 103d In- 
fantry attempted to advance the eight 

'See below, pp. 149-52 



hundred yards from Ilangana to Terere. 

Strong combat patrols went out in the 
morning of 26 July to fix the location 
of the Japanese pillboxes as accurately 
as possible. After their return, the artil- 
lery began firing at 1115, one hour be- 
fore the infantry was to attack. At 1 145 
the load's front was covered with smoke 
and under its cover the front-line com- 
panies withdrew a hundred yards. At 
noon the artillery put its fire on the Jap- 
anese positions directly in front. As the 
tanks were not quite ready at H Hour, 
1215, the artillery kept firing for ten 
more minutes. It lifted fire one hundred 
yards at 1225, and the 103d started for- 
ward. The tanks led the advance in the 
center; behind them was the infantry. 
Attached to the 103d for the attack 
were 2d Lt. James F. Olds, Jr., the act- 
ing corps chemical officer, and six volun- 
teers from the 11 8th Engineer Battalion. 
Each carried a flame thrower, a weapon 
which the 43d Division had brought to 
New Georgia but had not used up to 
now. 9 Griswold, whose headquarters had 
conducted flame thrower schools on 
Guadalcanal, was aware of the weapon's 
possibilities. That morning the six en- 
gineers had received one hour of train- 
ing in the use of the M1A1 flame 

The flame throwers went forward with 
the infantry, which halted about tw T enty 
yards in front of the pillbox line and 
covered it with small arms fire. Under 
cover of this fire the flame thrower op- 
erators, their faces camouflaged with dirt, 
crawled forward. Operating in teams of 

6 On ig July Griswold radioed Guadalcanal to 
state his urgent need for more M1A1 flame throw- 

two and three, they sprayed flame over 
three barely visible pillboxes in front of 
the center of the load's line. Vegetation 
was instantly burned off. In sixty seconds 
the three pillboxes were knocked out 
and their four occupants were dead. 10 

Operations of the infantry, tanks, 
flame throwers, and supporting heavy 
weapons and artillery met with almost 
complete success. The 103d Infantry en- 
countered seventy-four pillboxes on a 
600-yard front, but by midafternoon, 
spurred on by pressure from General 
Wing, it had reduced enemy resistance 
at Ilangana. From there it continued its 
advance through underbrush and vines 
and gained almost 800 yards. By 1700 
the left flank rested on the coastal village 
of Kia. The 43d Division's line, formerly 
1,700 yards long, was now much 
straighter by 300 yards. 

From 28 through 31 July, the 43d 
Division inched slowly forward, a few 
yards on the right flank and about five 
hundred yards along the coast. This was 
accomplished by "aggressive action and 
small unit maneuver, combined with 
constant artillery and mortar action 
[which] gradually forced the enemy back 
from his high ground defenses." 11 The 
i72d ground its way over Shimizu Hill, 
the last real ridge between it and Munda 
airfield, and in doing so it helped un- 
hinge the main Japanese defense sys- 
tem in its zone, just as the io3d's drive 

10 Capt. James F. Olds, Jr., "Flamethrowers Front 
and Center," Chemical Warfare Bulletin, Vol. 30, 
No. 3 (June-July 1944), pp. 5-9. This account, 
while valuable, seems to have telescoped two situ- 
ations and actions into one, for Olds asserts that 
the 103d was at Lambeti Plantation on 26 July. 
From the fact that three pillboxes had only four 
occupants, it would seem that this part of the 
Japanese line was lightly manned. 

"43d Div Rpt of Opns, Munda Campaign, p. 13. 



through Ilangana had broken the' enemy 
line on the left. 12 

Major Zimmer's ist Battalion, 169th 
Infantry, was brought over from Ren- 
dova on 29 July; the 3d Battalion, now 
commanded by Maj. Ignatius M. Ram- 
sey, was taken out of division reserve 
and the 169th (less the 2d Battalion) was 
assigned a zone between the i72d and 
the 103d. 13 As the month ended the 
169th (less its 2d Battalion in corps re- 
serve) was in the process of extending 
to the northwest to pinch out the 173d. 

Command of the 43d Division changed 
hands on 29 July when Maj. Gen. John 
R. Hodge, the tough, blunt commander 
of the Americal Division, came up from 
the Fijis to take over from Hester. This 
change was ordered by General Harmon 
who felt that Hester had exhausted him- 
self. General Hodge had served as assist- 
ant commander of the 25th Division dur- 
ing the Guadalcanal Campaign, and 
thus had had more experience in jungle 
warfare than any other general then in 
New Georgia. Hodge, Harmon wrote, 
was the "best Div Comdr I have in area 
for this particular job." 14 

The 43d Division, having cracked 
through the Shimizu Hill-Ilangana posi- 
tions, was in a favorable position to drive 

12 Unfortunately the records are too scanty to pro- 
vide details showing just how the 173d took this 
position. During the attack 1st Lt. Robert S. Scott 
almost singlehandedly halted a Japanese counter- 
attack and for his gallantry was awarded the Medal 
of Honor. WD GO 81, 14 Oct 44. 

13 Colonel Reincke was now regimental executive 

" See Ltr, Harmon to Griswold, quoted in 
SOl'ACBACOM, History of the New Georgia Cam- 
paign, I, 25, OCMH. See also Halsey and Bryan, 
Admiral Hahey's Story, p. 161; Rad 2027, COM- 
GENSOPAC to CofS USA, 10 Aug 43, in Marshall's 
IN Log. 

against Munda under its new com- 
mander, while the 37th Division on its 
right fought its way through the enemy 
positions in its hilly, jungled zone. 

The Attack Against the Ridges: 
The 37th Division 

The dawn of D Day, 25 July, found 
the 43d Division committed to a general 
attack, but the 37th Division was forced 
to postpone its advance. General Beight- 
ler had issued a field order on 23 July 
calling for a general attack by his three 
regiments, to start at 0700, 25 July. The 
145th, 161st, and 148th Infantry Regi- 
ments were to attack due west, toward 
Bibilo Hill, on the division's left, center, 
and right, with the 145th maintaining 
contact with the i72d Infantry on its 
left and the 148th Infantry covering the 
corps' right flank and rear. But the dis- 
covery of the strong Japanese position 
east of the 161st Infantry's line of de- 
parture altered the plans. 

On 24 July Beightler ordered Colonel 
Holland not to advance his 145th In- 
fantry, but to stay in place. He told 
Colonel Baxter to move the 148th In- 
fantry only up to the line of departure. 
The two regiments would hold while 
part of the iGist contained the Japanese 
position and the rest of the regiment 
bypassed it and came up on a line with 
the 145th and 148th. 

After Baxter received the commanding 
general's orders, he suggested that his 
regiment could perhaps help the 161st 
reduce the pocket by making a limited 
advance. Baxter hoped to establish an 
observation post on high ground from 
which both Munda airfield and the 
pocket in front of the iGist could be 



seen, Beightler assented to this request 
at 0910, 25 July. Patrols went out, and 
on their return the direct support artil- 
lery battalion laid a ten-minute prepara- 
tion 400 yards in front of the line of 
departure while mortars covered the 
400-yard gap. The 2d Battalion, 148th, 
commanded by Lt. Col. Herbert Rad- 
cliffe, started forward, met no Japanese, 
and gained 500-600 yards. 15 The 1st 
Battalion, Lt. Col. Vernor E. Hydaker 
commanding, moved up to the 2d Bat- 
talion's old positions. 

Bartley Ridge and Horseshoe Hill 

When I Company, 161st Infantry, had 
been unable to reduce the Japanese 
strongpoint on Bartley Ridge on 24 July, 
Colonel Dalton issued orders for its 
seizure on D Day. 10 I Company was to 
contain the Japanese pocket by attack- 
ing to its front while the 1st Battalion 
and the rest of the 3d Battalion exe- 
cuted a double envelopment. 17 The 1st 
Battalion was to move around the Jap- 
anese left (north) flank while the 3d Bat- 
talion went around the right, after which 
the two battalions would drive south- 
ward and northward for two hundred 
yards. Fifteen minutes of mortar fire 
would precede these moves. Beightler 
arranged for the 145th and 148th Regi- 
ments to support the 161st with heavy 

15 Because of the inaccurate maps and densely 
jungled, rough ground, all units had difficulty in 
determining distances and exact locations. Nearly 
all reports warn that distances and locations are 
only estimates, and all give widely varying figures. 

10 Bartley Ridge was named in memory of 2d Lt. 
Martin E. Bartley, an I Company platoon leader 
killed on 25 July. 

"The ad Battalion was in division reserve along 
with the 37th Cavalry Reconnaissance Troop. 

weapons fire. He also asked corps head- 
quarters for tanks to help the 161st, but 
the 43d Division had been given the 
tanks for the D-Day attack. 

From positions near the Laiana Trail 
eight 81-mm. mortars opened fire at 
0745, 25 July, in support of Dalton's at- 
tack. Heavy weapons of the adjoining 
regiments attempted to deliver their sup- 
porting fires, but the denseness of the 
jungle prevented forward observers' con- 
trolling the fire. The unobserved fire 
began obstructing rather than helping 
the 161st, and that part of the plan was 

The 3d Battalion, commanded by Lt. 
Col. David H. Buchanan, was unable to 
gain. Shortly after 0800, when the attack 
began, I Company reported that its at- 
tack against the ridge strongpoint had 
stalled. A knob projecting east from Bart- 
ley Ridge and the heavy undergrowth 
provided enough cover and concealment 
to let the infantrymen reach the base of 
the ridge, but uphill from the knob, 
where the growth was thinner, all move- 
ment was halted by fire from the crest. 
The 161st Infantry had made plans to 
use flame throwers, and an operator 
carrying his sixty-five pounds of equip- 
ment made two laborious climbs and 
silenced an enemy machine gun, but 
many other Japanese positions remained 
in action. The main body of the 3d 
Battalion, attempting to get around the 
south end of Bartley Ridge, was also 

The 1st Battalion was more successful. 
At 1035 its commander, Lt. Col. Slaft- 
cho Katsarsky, radioed Dalton that he 
had found the north flank of the Jap- 
anese position on Bartley Ridge, and 



that he was moving his battalion around 
it. Shortly afterward Beightler, Dalton, 
and staff officers conferred and decided 
that the 3d Battalion should contain 
the strongpoint while the 1st Battalion 
pushed westward with orders to develop 
enemy positions but not to engage in 
full-scale combat. The 37th Division 
could not advance westward in force 
until Bartley Ridge had been cleared. 

The 3d Battalion established itself in 
containing positions north, east, and 
south of Bartley Ridge. E Company was 
released from reserve and sent into line 
on high ground just north of Bartley to 
secure the right flank in the i6ist's zone. 
The 1st Battalion advanced to a point 
about four hundred yards west of Bart- 
ley and halted on a small rise northeast 
of Horseshoe Hill. Tanks of the newly 
arrived 10th Marine Defense Battalion 
were to be committed to support the 
37th Division the next day, and in the 
afternoon the tank commander made a 
personal reconnaissance of Colonel Bu- 
chanan's zone in preparation for the 

Six light Marine tanks were to lead 
out in the attack at 0900, 26 July, after 
preparatory fire by machine guns and 
81-mm. mortars. L and K Companies, in 
column, would move behind the tanks, 
which were supported by infantrymen 
armed with .30-caliber Mi rifles, .30- 
caliber Browning Automatic Rifles 
(BAR's), and two flame throwers. Tank- 
infantry communication was indirect. 
The tank radios formed a net within 
the Tank Platoon, and a 161st radio 
car maintained radio contact between 
the Tank Platoon commander and Col- 
onel Buchanan. 

When the tanks, with their hatches 

closed, got off the approach trail that 
had been bulldozed by members of the 
65th Engineer Battalion, Colonel Bu- 
chanan directed the infantrymen to lead 
them forward. It was 0925 before the at- 
tack got started. In two lines of three 
vehicles each, the tanks lumbered over 
the littered undergrowth, steep slopes, 
and felled logs toward the southeast 
slope of Bartley Ridge. The Japanese 
quickly responded with fire from anti- 
tank and 70-mm. battalion guns, ma- 
chine guns, and mortars. 

The attack went well at first. About 
a dozen pillboxes were reported knocked 
out by 1 no, and Buchanan ordered his 
men to occupy them to keep the Jap- 
anese from moving in again at night. 
Unfortunately, the tanks had encoun- 
tered exactly the sort of difficulties that 
might be expected in tangled terrain 
with communications uncertain. In their 
lurches and frequent changes of direc- 
tion they injured some of the accom- 
panying foot troops. Poor visibility 
caused them to get into untenable posi- 
tions from which they had to be extri- 
cated with consequent delays to the at- 
tack. During the morning a Japanese 
soldier stole out of the tangled brush 
and planted a magnetic mine that dis- 
abled one tank. A second tank was halted 
by a ruptured fuel line. The remaining 
four withdrew at 1110 to reorganize. 

The flame thrower operators, carry- 
ing their bulky, heavy fuel tanks on their 
backs, were not properly protected by 
the riflemen and were soon killed. 

In the course of the day's fighting some 
fourteen pillboxes and a number of ma- 
chine gun positions were knocked out, 
and the 3d Battalion advanced about 



two hundred yards up Bartley Ridge.' 8 
But it met such heavy fire from Bartley 
and Horseshoe Hill that its position 
clearly could not be held. Attempts to 
pull out the disabled tanks were unsuc- 
cessful. The battalion withdrew to its 
previous positions. The attack had dis- 
closed the existence of so many more 
positions that Dalton received Beight- 
ler's permission to make a thorough re- 
connaissance before attacking again. 

While the 1 6 1 st was attacking Bartley 
Ridge on 25 and 26 July, Colonel Hol- 
land's 145th Infantry stayed in its for- 
ward positions on Reincke Ridge and 
Kelley Hill. During this period it sent 
out patrols to the north and west to try 
to find the source of the 90-mm. mortar 
fire that had been hitting the regiment 
since 22 July. It received no artillery 
support at this time because its front 
line was considered too close to enemy 
targets for the artillery to fire without 
hitting American infantry. By the end 
of 26 July the 1st Battalion, 161st, had 
fought its way forward to come up on 
line north of the 145th, but near the 
regimental boundary the line sagged 
eastward in the shape of a great U. Col- 
onel Parker's 2d Battalion, 145th, was 
occupying positions in rear of the 1st 
Battalion, 145th, commanded by Lt. Col. 
Richard D. Crooks. 

General Beightler ordered Holland to 
commit his 2d Battalion, in order to 
reduce the Japanese positions on Horse- 
shoe Hill that had fired on the 3d Bat- 

a In the course of the day's action Capt. Paul K. 
Mellichamp, battalion executive officer, picked up 
a radio from a wounded operator and directed 
mortar fire. He was wounded, but continued to 
direct the fire until he collapsed. He died shortly 
afterward, and was awarded the Distinguished Serv- 
ice Cross. 

talion, 161st, during its 26 July attack 
against Bartley Ridge. Doubtless because 
troops of the 161st had not been able 
to get past the south end of Bartley, 
Colonel Parker's battalion was to march 
northward right around the 3d Battal- 
ion, 161st, push west around the north 
of the enemy positions on Bartley Ridge, 
then attack to the southwest. This ma- 
neuver would entail a march of about 
one and one-half miles to the assembly 

Parker's battalion moved out in the 
early morning of 27 July. It reached its 
assembly area on the north flank of 
Bartley Ridge without incident. After 
a preparation of one hundred rounds by 
the division artillery, which cleared some 
of the foliage, the battalion advanced to 
the attack in column of companies. 
"Having to fight every foot of the way," 
it gained about three hundred yards 
before 1300, when E Company in the 
lead moved south off slopes of a ridge 
and started up a small knob projecting 
from Horseshoe Hill. 19 As the company 
ascended the hill it was struck by fire 
from pillboxes. Among the first men 
killed was Capt. Gardner B. Wing, E 
Company's commander, in whose honor 
the 145th christened the knob. 

On the same day, while American 
mortars fired intermittently at Bartley 
Ridge, patrols from the 3d Battalion, 
161st Infantry, examined the Japanese 
lines to procure data for a preparation 
by the corps artillery. In the course of 
the reconnaissance Colonels Dalton and 
Buchanan observed enemy pillboxes on 
the right flank of the 1st Battalion, 145th 

la 37th Div G-3 Jnl, 27 Jul 43. 



Infantry, and recommended that the at- 
tack be delayed until 29 July. 20 General 
Beightler gave his assent. Dalton and 
Buchanan also decided to attack from 
the northwest instead of the southeast. 
Reconnaissance and pressure were to 
continue on the 28th. 

On the morning of the 28th a ten- 
man patrol from I Company, led by Lt. 
Walter Tymniak, set out in a southerly 
direction toward the top of Bartley 
Ridge. To their surprise and satisfac- 
tion, the Americans met no fire, got 
safely to the top, and found several 
abandoned pillboxes. They occupied 
them, and I Company followed to the 
crest and began infiltrating the pillboxes. 
Not all were vacant, but the task of the 
attackers was eased as each pillbox was 
taken, for its fire could then no longer 
be used with that of its neighbors to 
make crossfire or interlocking bands of 
fire. Because the Japanese appeared to 
be evacuating and the American front 
was intermingled with the enemy front, 
the artillery preparation was called off. 
The 3d Battalion continued its infiltra- 
tion on 29 July. At the end of the day 
it was relieved by Maj. Francis P. Car- 
berry's 2d Battalion and went into divi- 
sion reserve. 

The 145th Infantry's zone was shifted 
farther north on 30 July as part of a 
general shift in boundaries that General 
Griswold was making in order to widen 
the 43d Division's front. This move 
placed the southern half of Bartley Ridge 
within the i45th's zone. Colonel Park- 
er's 2d Battalion, 145th, had just com- 
pleted its move around the i6ist's north 

ao It is not clear whether these pillboxes were on 
Horseshoe Hill or were on the saddle connecting 
Bartley Ridge with Horseshoe Hill. 

flank, thence southwest against Horse- 
shoe Hill. On 30 July it was attached 
to the 161st for the completion of the 
reduction of Bartley Ridge and Horse- 
shoe Hill. 

Carberry's and Parker's battalions 
pushed their attacks on 30 July. In con- 
trast with Carberry's battalion, which 
met little resistance, Parker's men en- 
gaged in sharp fighting in the west. The 
Japanese who had evacuated the position 
facing Carberry had apparently moved 
into positions facing Parker. With gre- 
nade, rifle, machine gun, mortar, and 
flame throw T er the two battalions fought 
all day and part of the next, until by 
midafternoon of 31 July the Japanese 
rear guards on Bartley Ridge were either 
dead or in flight, and the 2d Battalion, 
161st, had advanced west and was on a 
line with the 1st. Bartley Ridge had con- 
tained forty-six log and coral pillboxes 
and thirty-two other lighter positions. 
First attacked by a company, it fell only 
after seven days' fighting by two battal- 

On Horseshoe Hill the Japanese re- 
sisted from their pillboxes and foxholes 
with equal skill and enthusiasm. The 
Americans used small arms, grenades, 
automatic weapons, mortars, flame 
throwers, and field artillery as they sys- 
tematically reduced the enemy positions, 
almost pillbox by pillbox. 21 On 1 August 
Parker's battalion received orders to at- 
tack in late afternoon, obeyed, and took 
Horseshoe Hill without firing a shot or 
losing a man. The Japanese had gone. 

41 During these operations Pfc. Frank J. Petrarca, 
a medical aid man, so distinguished himself by 
gallant, selfless devotion to duty that he was post- 
humously awarded the Medal of Honor. WD GO 
86, 23 Dec 43. 



Advance and Withdrawal 
of the 148th Infantry 

On 1 August, the day on which the 
Americans completely occupied the 
ridge positions, the 148th Infantry re- 
turned eastward to the 37th Division's 
lines after an advance which had taken 
it almost to Bibilo Hill. The 148th In- 
fantry was the only regiment not con- 
fronted by prepared enemy positions, 
and it had made comparatively rapid 
progress from the first. When Colonel 
Baxter moved his regiment forward on 

25 July, it went around the north flank 
of the Japanese defense line and met no 
resistance. However, none of the Ameri- 
cans then knew that the major part of 
the enemy 13 th Infantry lay to the north 
of Baxter's right flank. Patrols, accom- 
panied by Fiji scouts, went out and 
reported the presence of a few Japanese 
to the west, none to the south. Generals 
Griswold and Beightler had emphasized 
the importance of maintaining lateral 
contact and Beightler had expressly di- 
rected that the 148th was to maintain 
contact with the 161st, and that all units 
were to inform their neighbors and the 
next higher unit of their locations. The 
148th, however, was not able to make 
contact on its left with the 161st Infan- 

Baxter's two-battalion regiment ad- 
vanced regularly for the next three days. 
Colonel Radcliffe's 2d Battalion led on 

26 and 27 July; on 28 July Colonel Hy- 
daker's 1st Battalion bypassed the 2d 
and led the advance to a point some- 
where east of Bibilo Hill. 22 Patrols went 

M The total of daily yardage reported in the 
journals, if correct, would have placed the 148th 
west of Bibilo Hill on a8 July, but the 148th 
soldiers, like almost everyone else in the jungle, 
overestimated the distances they had traveled. 

out regularly and at no time reported 
the presence of a sizable body of the 
enemy. On 27 July Baxter reported that 
he had established "contact with Whisk- 
ers." Colonel Dalton, the "guest artist" 
regimental commander of the attached 
161st Infantry, sported a beard and was 
dubbed "Whiskers" and "Goatbeard" 
in the 37th Division's telephone code. 
But the i48th's front was almost a thou- 
sand yards west of Whiskers' 1st Battal- 
ion, and the contact must have been 
tenuous. Next day G Company was 
ordered to move to the left to close a gap 
between the two regiments, but the gap 
stayed open. 

During the move troops of the 117th 
Engineer Battalion labored to push a 
supply trail behind the advancing battal- 
ions. The rate of march was in part 
geared to the construction of the supply 
trail. As Baxter told Radcliffe over the 
telephone on 27 July, "I am advancing 
behind you as fast as bulldozer goes." 23 

Next day, however, there occurred a 
disturbing event. A platoon from A 
Company, 117th Engineer Battalion, 
was using a bulldozer to build the trail 
somewhere north of Horseshoe Hill 
when it was ambushed by the enemy. 
Three engineers were killed and two 
were wounded before elements of the 
Antitank Company and of the 1st Bat- 
talion rescued the platoon and extricated 
the bulldozer. 

Japanese movements during this pe- 
riod are obscure, but this and subse- 
quent attacks were made by the 13th In- 
fantry coming south at last in accordance 
with Sasaki's orders. 

The situation became more serious on 
28 July, the day on which Baxter's ag- 

a 148th Inf Jnl, S7 Jul 43. 



gressive movement took him almost to 
Bibilo Hill. At this time the regiment 
was spread thinly about fifteen hundred 
yards beyond the 161st; its front lay some 
twelve hundred yards west of the regi- 
mental ration dump and eighteen hun- 
dred yards from the point on the supply 
line "which could be said to be ade- 
quately secured by other division 
units." 24 There was still no contact with 
the 161st, and in the afternoon a group 
of the ijth Infantry fell upon the ration 
dump. From high ground commanding 
it the enemy fired with machine guns, 
rifles, and grenade discharges at men of 
the regimental Service Company. The 
Service Company soldiers took cover 
among ration and ammunition boxes 
and returned the fire. The dump, under 
command of Maj. Frank Hipp, 148th 
S-4, held out until relieved by two 
squads of the Antitank Company and 
one platoon from F Company. East of 
the dump, troops of the 13th Infantry 
also forced the i48th's supply trucks 
to turn back. Baxter, stating "I now find 
my CP in the front line," asked Beight- 
ler to use divisional units to guard the 
trail up to the dump. 25 

All the i48th's troubles with the 
Japanese were in the rear areas. The 
westward push, which took the leading 
battalion as far as one of the Munda— 
Bairoko trails, had been practically un- 
opposed. But early on the morning of 
29 July General Beightler, unaware of 
the ijth's position, telephoned Baxter to 
say that as the Japanese seemed to be 
moving from the southwest through the 
gap between the 148th and 161st Regi- 

2 ' 37th Div Opn Rpt, p. 6. 
25 37th Div G-3 Jnl, 2 8 Jul 43. 

ments, and around the i48th's right, 
Baxter was to close up his battalions and 
consolidate his positions. At 0710 Beight- 
ler told Baxter to withdraw his battal- 
ions to the east, to establish contact with 
the 161st, and to protect his supply route. 
Baxter, who had sent patrols out in all 
directions early in the morning, at 0800 
ordered one company of the 2d Battal- 
ion to clear out the supply trail to the 
east. At 0941, with Japanese machine 
guns still dominating the supply trail, 
Beightler sent Baxter more orders simi- 
lar to those of 0710, and also ordered 
forward a detachment of the 37th Cav- 
alry Reconnaissance Troop to help clear 
the east end of the supply trail. The 
telephone, so busy with conversations 
between Beightler and Baxter on 29 
July, was then quiet for an hour. 

Meanwhile Beightler had been con- 
ferring with Dalton, Holland, and mem- 
bers of the divisional general staff. As a 
result he had decided that the 161st 
should continue reducing Bartley Ridge, 
that the 145th should stay in place, and 
that the 148th would have to withdraw. 
So at 1055 Baxter ordered his regiment 
to turn around and pull back to the east. 
The 3d Battalion, 148th, was to use at 
least one company to establish contact 
with the 161st while the rest of the bat- 
talion withdrew toward the ration dump. 
The 1st Battalion would move back to 
the 2d Battalion's positions. At 1150 
Baxter reported the 2d Battalion in con- 
tact with the 161st, and shortly after- 
ward Beightler ordered Baxter to move 
the 1st Battalion farther east, putting 
it in position to deliver an attack the 
next morning against the rear of the 
Japanese holding up Dalton's regiment. 
The division commander again empha- 



sized the necessity for maintaining firm 
contact with the 16 1st. 26 At 1305, with 
the 148th moving east, Colonel Katsar- 
sky reported that his 1st Battalion, 161st, 
had as yet no contact with the 148th. 
Beightler at once told Baxter that, as 
Japanese machine gunners were operat- 
ing between the two regiments, the 
gap must be closed before dark. An hour 
later Baxter called Beightler to say that 
he was too far west to close with the 
161st before dark. When Beightler or- 
dered him to close up anyway, Baxter 
demurred. Asking his general to recon- 
sider the order, he stated that he could 
almost, but not quite, close the gap. 
Beightler thereupon told Baxter to com- 
ply with his orders as far as was physi- 
cally possible. 

The 2d Battalion had meanwhile been 
pushing east, except for F Company's 
main body, which was advancing w T est 
toward the ration dump. Both bodies 
were encountering enemy resistance, 
and the day ended before the Japanese 
were cleared out. The Reconnaissance 
Troop cleared some Japanese from the 
eastern part of the supply trail, but at 
1758 Baxter reported that the trail had 
been closed by raiding Japanese. 

The 148th Infantry, in examining the 
personal effects of some of the dead 
Japanese., found that the men belonged 
to the 13th Infantry. Some of them had 
been carrying booty taken in the raids 
east of the Barike several days earlier. 
Colonel Baxter later estimated that the 
enemy harrying his regiment numbered 
no more than 250, operating "in multi- 

18 As the Americans still did not know the 13th 
Infantry's location, they thought the attack had 
originated from the southwest rather than from 
the north. 

pie small light machine gun and mortar 
detachments and . . . [moving] from 
position to position utilizing the jungle 
to its maximum advantage. You can well 
imagine what we could do with our 
M i's, BAR's and Machine Guns if all 
we had to do was dig in and wait for 
the Jap to come at us." 27 

General Beightler, a National Guards- 
man most of his life, was an affable man, 
but he was far from satisfied with the 
outcome of the day's action. 28 At 1832 
he radioed Baxter that General Gris- 
wold had ordered the 148th to establish 
contact with the 161st early the next 
morning and to protect the supply route. 
"Use an entire battalion to accomplish 
latter if necessary. At no time have you 
been in contact on your left although 
you have repeatedly assured me that 
this was accomplished . . . Confirma- 
tion of thorough understanding of this 
order desired.'' Baxter thereupon tele- 
phoned division headquarters and put 
his case before a staff officer. General 
Beightler's criticism, he felt, was not 
justified. "Please attempt to explain to 
the General that I have had patrols 
in contact with the 161 and have docu- 
mentary evidence to substantiate this. 
I have not, however, been able to 
maintain contact and close the gap 
by actual physical contact due to the 
fact that the 161st had been echeloned 
600 to 800 yards to my left rear. I have 
been trying and will continue tomorrow 

" Opns of the 148th Inf in New Georgia, 18 Jul- 
5 Aug 43, p. 8i. 

M General Beightler's record as a division com- 
mander was somewhat unusual in that he was in 
command of the 37th at the outbreak of war, led 
it overseas in 1942, and retained command of it 
through the Solomons and Philippines campaigns 
until the end of the war. 



morning to establish this contact. It is 
a difficult problem as I have had Japs be- 
tween my left flank and the 161st." 29 

Rain and mud added to Baxter's trou- 
bles on 30 July. Still harried by enemy 
machine guns and mortars, the 2d Bat- 
talion pushed east and south toward the 
161st as the 1st Battalion covered the left 
(north) flank. Elements of the 37th Cav- 
alry Reconnaissance Troop, C Company, 
117th Engineer Battalion, and the 3d 
Battalion, 161st Infantry, pushed north 
to give additional protection to the di- 
vision's right (north) flank and to pro- 
tect the east end of Baxter's supply route. 
Baxter attempted to cut a new trail di- 
rectly into the i6ist's lines, but Japanese 
rifle fire forced the bulldozer back. Some 
of the i48th's advance elements side- 
slipped to the south and got through 
to the 161st that day, but the main 
body was still cut off. 30 Some of the Jap- 
anese who were following the 148th at- 
tacked the 1st Battalion, 161st, but 
were halted. This action then settled 
down into a nocturnal fire fight. 

The plight of the rest of the regiment 
was still serious. Water was running 
low. Part of the Reconnaissance Troop 
tried to take water forward to the 148th 
on 30 July. It was stopped by Japanese 
fire. But rain fell throughout the night 
of 30-31 July and the thirsty men were 
able to catch it in helmets and fill their 

On 31 July Beightler suggested that 

28 The version of the radiogram quoted is taken 
from the 37th Div G-g Jnl File, ag Jul 43; the tele- 
phone message is from the 37th Div G-3 Jnl, 29 
Jul 43. 

:m F and H Companies, part of E Company, and 
the 2d Battalion Headquarters Company were the 
elements that got through to the 161st. 

Baxter destroy heavy equipment and 
break his regiment into small groups to 
slip northward through the jungle 
around the enemy. The 148th blew up 
all the supplies it could not carry but 
it had to fight its way along the trail. It 
had over a hundred wounded men and 
could not infiltrate through the jungle 
without abandoning them. 

Toward the end of the day B Com- 
pany, which had been trying to clear 
the Japanese north of the supply trail, 
was ordered to disengage and withdraw 
slightly for the night. One of B Com- 
pany's platoons, however, had come un- 
der fire from a Japanese machine gun 
about seventy-five yards to its front and 
found that it could not safely move. 
Pvt. Rodger Young, who had been 
wounded in the shoulder at the first 
attempt to withdraw, told his platoon 
leader that he could see the enemy gun 
and started forward. Although a burst 
from the gun wounded him again and 
damaged his rifle, he kept crawling for- 
ward until he was within a few yards 
of the enemy weapon. As a grenade 
left his hand he was killed by a burst 
that struck him in the head. But he had 
gotten his grenade away, and it killed 
the Japanese gun crew. His platoon was 
able to withdraw in safety. For his gal- 
lantry and self-sacrifice Young was post- 
humously awarded the Medal of Honor. 31 

Colonel Baxter's radio fairly crackled 
the next morning, 1 August, with orders 
from General Beightler: "Time is pre- 
cious, you must move." "Get going." 

31 Stanley A. Frankel, The 37th Infantry Division 
in World War II (Washington: Infantry Journal 
Press, 1948), p. 101; WD GO 3, 6 Jan 44. Mr. Frank 
Loesser commemorated this exploit in a popular 
song, "Rodger Young," copyright 1945. 



"Haste essential." 32 Thus urged on, 
Baxter ordered an assault by every man 
who could carry a rifle. He formed all 
his command— A, E, B, and G Compa- 
nies—in a skirmish line with bayonets 
fixed, and assaulted by fire and move- 
ment at 0850. The attack succeeded. By 
0930 the leading elements, ragged, 
weary, and muddy, reached Katsarsky's 
area. The 148 th was given fresh water 
and hot food, then passed into division 
reserve. As the men struggled in after 
their ordeal, all available ambulances, 
trucks, and jeeps were rushed up to 
transport the 128 wounded men to the 
37th Division's clearing station at 

Capture of the Airfield 

The first day of August had broken 
bright and clear after a night of intermit- 
tent showers. It is likely that the spirits 
of the top commanders were also bright, 
for things were looking better. With 
Ilangana and Shimizu Hill reduced, the 
43d Division was in possession of the 
last piece of high ground between it and 
Munda airfield. Bartley Ridge had 
fallen; Horseshoe Hill was about to fall, 
and the 148th was completing its retire- 

General Griswold had issued no spec- 
ial orders for the day; the field order 
that had started the corps offensive was 
still in effect. In the 37th Division's zone 
the most significant development was 
the return of Baxter's men. The 145th 
Infantry was patrolling; the 161st was 
mopping up. In the 43d Division's area 
of responsibility, General Hodge had 
ordered an advance designed to bring 

his division up on line with the 145th 

The 103d Infantry began its attack 
at 1100. E, G, and F Companies ad- 
vanced in line behind patrols. Meeting 
practically no opposition, they gained 
ground rapidly and by 1500 were nearing 
Lambeti Plantation. The 2d Battalion, 
169th Infantry, then in process of pinch- 
ing out the i72d, attacked northwest 
across the front of the i72d and estab- 
lished contact with the 145th Infantry. 
The i72d completed a limited advance 
before going into division reserve. The 
3d Battalion, 169th, on the left of the 2d, 
attacked in its zone and at 1500 was still 
advancing. For the first time since it had 
landed on New Georgia, the 43d Division 
could announce that the going was easy. 

The day before, Generals Hodge and 
Wing, accompanied by Colonel Ross, 
had visited the command and observa- 
tion posts of the 1st Battalion, 145th In- 
fantry, from where they could see part 
of Munda airfield. They detected evi- 
dence of a Japanese withdrawal, which 
seemed to be covered by fire from the 
enemy still on Horseshoe Hill. 

Thus at 1500, 1 August, with the 43d 
Division still moving forward, General 
Griswold ordered all units to send out 
patrols immediately to discover whether 
the Japanese were withdrawing. "Smack 
Japs wherever found, even if late." 33 If 
the patrols found little resistance, a gen- 
eral advance would be undertaken in 
late afternoon. Colonel Ridings tele- 
phoned the orders to 37th Division 
headquarters, and within minutes pa- 
trols went out. They found no enemy. 
At 1624 Ridings called Beightler's head- 

37th Div G-3 Jnl, 1 Aug 43. 

XIV Corps G-3 Jnl, 1 Aug 43. 



quarters again with orders to advance 
aggressively until solid resistance was 
met, in which case its location, strength, 
and composition were to be developed. 
The 148th Infantry was to have been 
placed in corps reserve with orders to 
protect the right flank, patrol vigorously 
to the north, northeast, and northwest, 
and cut the Munda-Bairoko trail if pos- 
sible. Since the 27th Infantry of the 25th 
Division was arriving and moving into 
position on the 37th Division's right 
flank, Beightler persuaded corps head- 
quarters to let him use the 148th Infan- 
try. 34 Ridings required, however, that 
the 148th be given the mission of pro- 
tecting the right flank because the 27th 
Infantry would not have enough strength 
for a day or two. 

All went well for the rest of the day. 
The 103d Infantry reached the outer 
taxiways of Munda airfield; the 169th 
pulled up just short of Bibilo Hill. The 
37th Division's regiments plunged for- 
ward past Horseshoe Hill, which was 
free of Japanese, and gained almost 
seven hundred yards. 

The Japanese Withdrawal 

The Japanese positions facing the XIV 
Corps had been formidable, and the 
Americans had been held in place for 
long periods. But the Americans had 
wrought more destruction than they 
knew. The cumulative effect of continu- 
ous air and artillery bombardment and 
constant infantry action had done tre- 
mendous damage to Japanese installa- 
tions and caused large numbers of cas- 
ualties. By late July most of the Japanese 
emplacements near Munda were in 

M See below, pp. 1(17-69 

shambles. The front lines were crum- 
bling. Rifle companies, 160-170 men 
strong at the outset, were starkly re- 
duced. Some had only 20 men left at 
the end of July. The 229th Infantry num- 
bered only 1,245 effectives. Major Hara, 
Captain Kojima, and many staff officers 
of the 229th had been killed by artillery 
fire. Hospitals were not adequate to care 
for the wounded and sick. The constant 
shelling and bombing prevented men 
from sleeping and caused many nervous 

To compensate for the diminution of 
his regiment's strength, Colonel Hirata 
ordered the soldiers of his 229th Infan- 
try to kill ten Americans for each Jap- 
anese killed, and to fight until death. 

Higher headquarters, however, took 
a less romantic view of the situation. On 
29 July a staff officer from the 8th Fleet 
visited Sasaki's headquarters and or- 
dered him to withdraw to a line extend- 
ing from Munda Point northeast about 
g,8oo yards inland. The positions facing 
the XIV Corps, and Munda airfield it- 
self, were to be abandoned. Sasaki and 
his subordinates thought that it would 
be better to withdraw even farther, but 
the views of the 8th Fleet prevailed over 
those of the responsible men on the spot. 
The withdrawal, which was deduced by 
XIV Corps headquarters on 1 August, 
was accomplished promptly, and except 
for detachments at Munda and in the 
hills the main body of Sasaki's troops 
was in its new position by the first day 
of August. 

Jungle Techniques and Problems 

The Americans did not yet know it, 
but the worst was over. All regiments 
began making steady progress each day 



Munda Airfield 

its own as an offensive weapon. All regi- 
ments employed it against enemy posi- 
tions, both in assault and in mopping 
up. The flame thrower did have several 
important disadvantages. The equip- 
ment was large and heavy, and required 
the operator to get very close to enemy 
positions, then expose his head and body 
in order to use his weapon. He needed 
to be protected by several riflemen. But 
even with its disadvantages, it was useful 
in destroying enemy positions. 

Tanks, too, were of great value. Gen- 
eral Griswold felt that, despite the diffi- 
culties inherent in operations over hilly 
jungle, the actions of the Tank Platoons 

against light, though determined and 
skillful, opposition. 35 

By now alt regiments, though depleted 
by battle casualties and disease, had be- 
come veterans. Pockets that once would 
have halted an entire battalion or even 
a regiment were now usually reduced 
with speed and skill. The flame thrower, 
receiving its most extensive use in the 
Pacific up to this time, was coming into 

35 Unless otherwise indicated all lessons-learned 
data are from CG XIV Corps, Informal Report on 
Combat Operations in the New Georgia Campaign. 
This document is not a narrative report. It contains 
data on tactics, weapons, logistics, and special jungle 
problems as compiled by the corps, division, and 
special headquarters participating in the operation. 



of the gth and 10th Marine Defense Bat- 
talions had been successful. On 29 July, 
looking forward to fighting over easier 
terrain around Munda airfield, he asked 
General Harmon for more tanks. Corps 
headquarters, he also announced, was 
preparing to mount flame throwers on 
tanks.™ The operation ended before 
flame throwing tanks could be used, but 
the idea was successfully carried out in 
later campaigns. 

The technique of reducing a pillbox, 
whether isolated or part of a defensive 
system, was now mastered. The official 
records unfortunately do not give much 
exact information on the reduction of 
specific pillboxes, but after the battle 
the 37th Division gave a valuable general 
description of the methods employed. 
The first essential was a complete recon- 
naissance to develop the position, in- 
tention, and strength of the enemy. This 
was quite difficult in the jungle. "To 
one unskilled in jungle fighting, it is in- 
conceivable that well trained reconnais- 
sance patrols in sufficient numbers can- 
not develop the situation in front of the 
advancing forces." 37 Because they could 
not see far enough, because they could 
not always get close enough, and because 
Japanese fire discipline was sometimes 
so good that a given position would not 
fire until actually attacked, reconnais- 
sance patrols could not always develop 
positions. The next step was a recon- 

26 Rati, Griswold to Harmon, 29 Jul 43, in XIV 
Corps G-3 Jnl, 30 Jul 43. Part of the 754th Army 
Tank Battalion was alerted at Noumea for transfer 
to Guadalcanal to be equipped with flame throwers 
for employment in New Georgia but Munda airfield 
had fallen before it was moved. 

The Tank Platoon of the nth Marine Defense 
Battalion arrived in early August. 

" 37th Div Rpt, G— 3 Narrative, Jungle Tactics 
and Opns, p. g. 

naissance in force by a reinforced pla- 
toon. This often uncovered a portion 
of the enemy position but not all of it. 
Usually the complete extent of a center 
of resistance was determined only by 
the attack. 

The attack itself consisted of three 
parts: artillery preparation, 81 -mm. mor- 
tar fire, and assault. 

The artillery preparation had a three- 
fold effect. It improved visibility by 
clearing away brush and foliage. It de- 
stroyed or damaged enemy positions. 
And it killed, wounded, and demoralized 
enemy soldiers. 

The 81 -mm. mortars, using heavy shell 
that had a delay fuze, fired on observed 
positions and usually covered the area 
between the American infantry and the 
artillery's targets. They frequently drove 
the Japanese soldiers out of their pill- 
boxes into the open where they became 
targets for rifle and machine gun fire. 
The 60-mm. mortars, though more mo- 
bile than the 8i-mm.'s, threw too light 
a shell to be very effective in these at- 
tacks. Their shells usually burst in the 
trees, but the 81-mm. heavy shells pene- 
trated the treetops and often the tops of 
the pillboxes themselves before explod- 

The assault consisted of a holding at- 
tack by a company or platoon delivering 
assault fire to cover a close-in single or 
double envelopment. BAR's, Mi's, and 
grenades were used extensively, and 
flame throwers were employed whenever 
possible. Units of the 25th Division, 
which later drove northward from 
Munda to Zieta, encountered pillbox 
positions that were too shallow, and in 
country too dense, for artillery and mor- 
tars to be used without endangering the 



attacking infantry. Men of this division 
therefore advocated flame throwers, in- 
fantry cannon, and tanks for pillbox 

These techniques, which simply repre- 
sented the application of established tac- 
tical principles, were being applied well 
in early August, but several problems 
remained. Because the infantry units did 
not advance at the same rate, the front 
line became irregular and the supporting 
artillery was thus unable to capitalize 
on the advantages of firing at right angles 
to the axis of advance. All unit com- 
manders were eager to employ artillery 
and mortar support to the utmost, but 
they frequently complained that neigh- 
boring units' supporting artillery and 

mortar fires were falling in their areas 
and endangering their troops. They had 
a tendency to forget that the enemy also 
used artillery and mortars and, when 
receiving American artillery fire, fre- 
quently lobbed 90-mm. mortar shells 
into the American front lines to convince 
the American infantrymen that they 
were being fired on by their own artil- 
lery. In most cases the complaints were 
probably caused by Japanese rather than 
American fire. 

Because maps were inaccurate and re- 
connaissance was inhibited by poor vis- 
ibility, it was extremely difficult to de- 
termine the exact location of friendly 
units. In the 37th Division's zone several 
artillery preparations were called off be- 



cause of uncertainty about the position 
of the 148th Infantry. Flares and smoke 
pots, and sometimes flame throwers, were 
used to mark flanks, but usually could 
not be seen by anyone not in the imme- 
diate vicinity. Griswold had ordered the 
front line battalions to mark their flanks 
daily with white panels twenty-five feet 
long by six feet wide. These were to be 
photographed from the air. Reconnais- 
sance planes made daily photographic 
flights, but there were no clearings in 
the New Georgia jungle large enough 
to permit the panels to be spread out, 
and this effort failed. By plotting close-in 
defensive artillery fires, forward observ- 
ers were able to provide some reliable 
information on the location of front 
lines. When the 37th Division rolled 
forward after 1 August, it estimated po- 
sitions and distances on the basis of 
speedometer readings from locations that 
had been plotted by air photography 
and interpolated on maps. 

The difficulties of scouting and patrol- 
ling naturally affected nearly every as- 
pect of the operation. Because enemy 
positions could not be fixed in advance, 
the troops often attacked terrain rather 
than the enemy. This procedure resulted 
in slow advances and in a high expendi- 
ture of mortar ammunition on areas 
actually free of the enemy. And mortar 
ammunition supply was laborious; shells 
had to be hand-carried from trail-end to 
the mortar positions. Poor scouting 
caused battalions to advance on narrow 
fronts and thus be halted by small en- 
emy positions. One regimental opera- 
tions officer asserted that inadequate re- 
connaissance was due in part to the fact 
that "higher commanders" did not issue 
orders until the late afternoon preced- 

ing an attack. Thus battalions did not 
have time for full reconnaissance: 

"Many times, units were committed 
in an area which had not been recon- 
noitered. This fact resulted in command- 
ers having to make decisions concerning 
a zone of advance in which he knew little 
or nothing about the enemy positions. 
Enemy strong points encountered in this 
fashion often times resulted in hasty 
withdrawals which were costly both in 
men and weapons." 38 

"Munda is yours" 

The XIV Corps maintained the mo- 
mentum of its advance against the en- 
emy delaying forces. On 2, 3, 4, and 5 
August the advance continued all across 
the corps' front. The 103d and i6gth 
Infantry Regiments, which had gained 
the outer taxiways of the airfield on 1 
August, kept going. The 3d Battalion, 
i72d, was committed on the i6gth's 
right on 4 August. In the more open 
terrain around the airstrip the troops 
were able to use 60-mm. mortars effec- 
tively, and their advance was conse- 
quently speeded. Kokengolo Hill, the 
rise in the center of the airfield where 
a Methodist mission had once stood, 
held up the advance temporarily. Bibilo 
Hill, whose fortifications included six 
75-mm. antiaircraft guns that the Jap- 
anese had been using as dual-purpose 
weapons, was reduced in three days of 
action by elements of the 169th, i72d, 
145th, and 161st Regiments, supported 
by Marine tanks. The 148th Infantry, 
on the north flank, established blocks 
and ambushes on a north— south track 

J8 Rpt, Maj Carl H. Coleman, S-3 145th Inf, to 
G-3 37th Div, 1 Sep 43, sub: Informal Rpt on New 
Georgia Campaign. 



that was presumed to be the Munda- 
Bairoko trail. 

On 5 August, with Bibilo Hill cleared, 
the units of the 37th Division crossed 
the narrow strip of land between the 
hill and the water. This tactical success 
had one effect of great personal impor- 
tance to the soldiers: many had their 
first bath in weeks. 

In the 43d Division's zone on 5 Aug- 
ust, the infantry, with tank and mortar 
support, killed or drove the last Jap- 
anese from the tunnels, bunkers, and 
pillboxes of Kokengolo Hill. Here were 
found caves stocked with rice, bales of 
clothing and blankets, and occupation 
currency. Crossing the western part of 
the runway, with its craters, grass, and 
wrecked Japanese planes, the infantry- 

men secured it in early afternoon. Gen- 
eral Wing telephoned General Hodge 
from Bibilo Hill: "Munda is yours at 
1410 today." 39 Griswold radioed the 
good news to Admiral Halsey: ". . . Our 
ground forces today wrested Munda 
from the Japs and present it to you . . . 
as the sole owner. ..." Halsey responded 
with "a custody receipt for Munda .... 
Keep 'em dying." 40 

The major objective was in Allied" 
hands. The hardest part of the long New 
Georgia battle was over. 

"•43d Div G-3 Jnl, 5 Aug 43. 

40 Rad, CG NGOF to COMSOPAC, 5 Aug 43, in 
XIV Corps G-3 Jnl, 5 Aug 43; Rad, COMSOPAC 
to CG NGOF, 6 Aug 43, in XIV Corps G-3 Jnl, 7 
Aug 43. 


After Munda 

The hardest slugging was over, at least 
on New Georgia. But several tasks faced 
the troops. The airfield had to be put 
into shape at once and the remaining 
Japanese had to be cleaned out of New 
Georg ia and sev eral of the offshore is- 

lands. (Map n) 

The Airfield 

Repair and enlargement of the bat- 
tered airstrip began immediately after 
its capture. 1 ". . . Munda airfield looked 
like a slash of white coral in a Dore 
drawing of hell. It lay like a dead thing, 
between the torn, coffee-colored hills 

1 Unless otherwise indicated this chapter is based 
on Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 237-44; Morison, Breaking the Bis- 
marcks Barrier, pp. 225-39; ONI USN, Combat 
Narratives: Solomon Islands Campaign, XI, Kolom- 
bangara and Vella Lavella, 6 August-7 October 
1943 [Washington, 1944]; SOPACBACOM, History 
of the New Georgia Campaign, Vol. I, Chs. VII- 
VIII; the rpts, jnls, and jnl files of NGOF, COMAIR 
New Georgia, XIV Corps, 25 th Div, 43d Div, 
northern landing force, and component units; His- 
tory of the 8th Area Army, 1942—44, Japanese 
Monogr No. 37 (OCMH); 17th Army Operations, 
Vol. II, Japanese Monogr No. 40 (OCMH); South- 
east Area Naval Operations, Vol, H, Japanese 
Monogr No. 49 (OCMH); outline of Southeast Area 
Naval Air Operations, August 1942-October 194a, 
Pt. II, Japanese Monogr No. 106 (OCMH). This 
section is also based on Building the Navy's Bases 
in World War II: History of the Bureau of Yards 
and Docks and the Civil Engineer Corps, 1940— 
1946, II (Washington, 1947), 265-66. 

of Bibilo and Kokengolo." 2 Seabees of 
the 73d and 24th Naval Construction 
Battalions began the work of widening, 
resurfacing, and regrading the field. On 
9 August additional naval construction 
battalions added their tools and men 
to the task. Power shovels dug coral out 
of Kokengolo Hill, and bulldozers, earth- 
movers, graders, and rollers spread and 
flattened it. Good coral was plentiful, 
as were men and tools, and the work 
moved rapidly forward. By 7 August 
the field, although rough, was suitable 
for emergency wheels-up landings. 

Advance parties from General Mul- 
cahy's air headquarters moved from Ren- 
dova to Munda during the second week 
of August. On the 14th, the day after 
the first Allied plane landed, General 
Mulcahy flew from Rendova to Munda 
in his amphibian plane and opened 
Headquarters, Air Command, New 
Georgia, in a Japanese-dug tunnel in 
Kokengolo Hill. 

Two Marine fighter squadrons (VMF 
123 and VMF 124), with twelve Cor- 
sairs (F4U's) each, arrived on the 14th 
and began operations at once. Together 
with the fighters based at Segi Point, 

'John A. DeChant, Devilbirds: The Story of 
United States Marine Corps Aviation in World War 
II (New York: Harper 8c Brothers, 1947), p. 109. 




which were also under Mulcahy, they 
and other Allied squadrons covered the 
Allied landing at Vella Lavella on 15 
August. 3 There were some difficulties at 
first. Maintenance crews were inexperi- 
enced, there were not enough spare 
parts, the field was not complete, and 
taxiways and dispersal areas were small 
and in poor condition. Japanese naval 
guns, promptly nicknamed "Pistol Pete," 
shelled the airfield from the nearby is- 
let of Baanga intermittently from 16 
through 19 August. But conditions 
quickly improved, and Pistol Pete, which 
had not done much damage, was cap- 
tured by elements of the 43d Division 
on 19 August. 

As the field was enlarged, more planes 
and units continued to arrive. Opera- 
tions intensified, and as the Japanese 
were cleared from the central Solomons 
Mulcahy's planes began to strike targets 
in the northern Solomons. For this rea- 
son his command was removed from the 
New Georgia Occupation Force on 23 
September and assigned as part of the 
Air Command, Solomons. Mulcahy's 
fighters escorted bombers to the Bou- 
gainville bases, and Munda-based bomb- 
ers soon began dropping loads there too. 

Munda airfield, which by mid-October 
had a 6,000-foot coral-covered runway 
and thus was suitable for bombers, be- 
came the best and most-used airfield in 
the Solomons. The rotation of Navy, 
Marine Corps, and Army Air Forces 
commanders that was standard in the 
South Pacific had brought about the 
relief of Admiral Mitscher as Com- 
mander, Aircraft, Solomons, by General 
Twining, the Thirteenth Air Force com- 
mander. General Twining moved his 

headquarters on 20 October from Guad- 
alcanal to Munda and made the most 
intensive possible use of the new base. 


Airfield development, though of pri- 
mary importance, could be of only minor 
interest to the ground troops who had 
the dreary task of slogging northward 
from Bibilo Hill in an attempt to trap 
the retreating Japanese. The job had 
been assigned to the 27th and 161st In- 
fantry Regiments, both operating under 
their parent command, the 25th Divi- 

Addition of the 27th Infantry to the 
New Georgia Occupation Force had 
come about because of General Gris- 
wold's urgent requests for more men. 
During July the Western Force of Task 
Force 31 had carried fully 26,748 men 
to Rendova, 4 but by the month's end 
not that many men were available for 
combat. Many of the arrivals were serv- 
ice troops. Further, casualties and dis- 
ease had weakened the infantry regi- 
ments. The three infantry regiments in 
the 37th Division (less the two battalions 
with Liversedge) had an authorized total 
strength of about 7,000 men. But the 
161st Infantry, which entered the cam- 
paign below strength, was short 1,350 
men. The two-battalion regiments were 
short too, so that the 37th Division's 
rifle regiments had only 5,200 men. And 
the 43d Division was in worse shape. 
With an authorized strength of 8,000 
men, its three rifle regiments had but 

'See below, pp. 179— 80 

4 CTF 31 War Diary, 31 Jul 43 entry. Also trans- 
ported were: 4,800 tons of rations; 17,431 drums of 
fuel, or 3,486 tons; 2,281 vehicles weighing 6,895 
tons; 9,961 tons of ammunition; and 5,323 tons of 
other freight. 



Munda Airfield in Operation. transport taking off is evacuating wounded men. 

4,536 men. Griswold had asked Liver- 
sedge if he could release two infantry 
battalions for the Munda drive but 
Liversedge replied that that would be 
possible only if Enogai and Rice Anchor- 
age were abandoned. His raider battal- 
ions were then only 60 percent effec- 
tive. 5 

On 28 July Griswold asked Harmon 
for replacements or for a regimental 
combat team less artillery. This request 
posed a grave problem for Harmon and 
Admiral Halsey. The injunction against 

3 See XIV Corps G-g Jnl File, ?8-?g, 31 Jul 43, 
By 14 August sickness and casualties had rendered 
the 4th Marine Raider Battalion practically unfit 
for fighting. The battalion surgeon, Lt. J. C. Lock- 
hart, USNR, reported that out of 453 men present 
only 137 were fit for duty. Memo, 4th Mar Raider 
Bn Surgeon for CO 4th Mar Raider Bn, 14 Aug 43, 
sub: Health of Personnel of 4th Raider Bn, in 
XIV Corps G-g Jnl File, 23 Aug 43. 

committing major forces to New Georgia 
was still in effect; at least it was theoreti- 
cally in effect, for in small island war- 
fare, especially in 1943, 26,000 men con- 
stituted a major force. The only imme- 
diately available division was the 25th, 
and one of its regiments, the 161st, had 
been sent in fairly early. Further, Halsey 
and Harmon had planned to use the 
25th for the invasion of the Buin-Faisi 
area of Bougainville. 

Yet as long as the high command re- 
tained confidence in Griswold, there 
could be but one answer. As Harmon 
wrote to his chief of staff, Brig. Gen. 
Allison J. Barnett, ". . . we have to 
make this Munda-Bairoko business go— 
and as quickly as possible. It is the job 
'in hand' and whatever we use we have 
to get it done before we go on to the 
next step." One of the major difficulties, 



according to Harmon, was the fact that 
the Americans had underestimated the 
job in hand. "Munda is a tough nut- 
much tougher in terrain, organization 
of the ground and determination of 
the Jap than we had thought. ... In 
both Guadalcanal and New Georgia we 
greatly underestimated the force re- 
quire to do the job." 6 Thus Harmon 
alerted the 27th Infantry of the 25th 
Division for transfer to New Georgia 
and recommended to Halsey that the 
25th Division be taken off the list for 
Bougainville. 7 As soon as he received 
Halsey's approval Harmon ordered up 
the 27th Infantry. On 29 July Col. 
Thomas D. Roberts of Harmon's staff 
arrived at Griswold's headquarters to 
announce the imminent arrival of the 
27th Infantry and some replacements. 

At this time the Japanese were still 
holding the Ilangana-Shimizu Hill- 
Horseshoe Hill-Bartley Ridge defense 
line, and no one was anticipating the 
rapid advances that characterized the first 
days of August. Thus on go July with 
Colonel Roberts' concurrence Griswold 
asked for more 25th Division troops and 
Harmon promptly promised the 35th In- 

Advance elements of the 27th Infan- 
try, and Headquarters, 25th Division, 
landed on the barrier island of Sasavele 
on 1 August, and in the next few days 
the regiment was moved to the right 
(north) flank of the Munda front to pro- 
tect the XIV Corps' right flank and rear. 

6 Ltr, Harmon to Harriett, 28 Jul 43, quoted in 
SOPACBACOM, History of the New Georgia Cam- 
paign, Vol. I, Ch. IV, p. 35, OCMH. 

7 He suggested substituting the 2d Marine Di- 
vision or the 3d New Zealand Division. 

The Cleanup 

North to Bairoko 

The Japanese withdrawal from Munda 
released a sizable body of American 
troops to attempt the cleanup of the 
Japanese between Munda and Dragons 
Peninsula. After the rapid advances be- 
gan on 1 August the 27 th Infantry, tem- 
porarily commanded by Lt. Col. George 
E. Bush, sent out patrols to the north 
before advancing to clear out the Jap- 
anese and make contact with Liversedge. 8 

Meanwhile General Griswold decided 
that mopping-up operations would have 
to include a drive from Bibilo Hill 
northwest to Zieta, a village on the west 
coast about four crow's-flight miles 
northwest of Bibilo, to cut off the re- 
treating Japanese. On 2 August the 37th 
Division had reported that Fijian pa- 
trols had cut the Munda-Bairoko trail 
but found no evidence of any Japanese 
traffic. Lt. Col. Demas L. Sears, Assistant 
Chief of Staff, G-2, 37th Division, of- 
fered the opinion that if the Japanese 
were evacuating New Georgia they were 
moving along the coast to Zieta rather 
than to Bairoko. This opinion was but- 
tressed by reports from Colonel Griffith 
of the Raiders who radioed on 2 August 
that there had been no traffic in or out 
of Bairoko. 

Next day, on orders from Griswold, 
the 3d Battalion, 148th Infantry, left 
Enogai on a cross-country trek toward 
Zieta, a trek that was halted short of 
there on 5 August by additional orders 

8 Col. Douglas Sugg had commanded the regi- 
ment until a few days before the move to New 
Georgia. He fell ill and was hospitalized, and his 
place was taken by Colonel Bush, the executive. 
Sugg resumed command of his regiment on 12 



4-T0N Truck Stuck in the Mud on a jeep trail, New Georgia. 

from Griswold. He had decided to use 
the two 25th Division regiments under 
General Collins, the commander who 
had led the division on Guadalcanal, to 
drive to Zieta and Bairoko. 

From then until 25 August, the 25th 
Division units slogged painfully along 
the swampy jungle trails in pursuit of 
the elusive enemy. The Japanese occa- 
sionally established roadblocks, am- 
bushes, and defenses in depth to delay 
the Americans, but the worst enemy was 
the jungle. The terrain was, if anything, 
worse than that encountered on the 
Munda front. The maps were incorrect, 
inexact, or both. For example, Mount 
Tirokiamba, a 1,500-foot eminence re- 
ported to lie about 9,000 yards north- 

west of Bibilo Hill, was found to be 
4,500 yards south by west of its reported 
position. Mount Bao, thought to be 
6,000 yards east-northeast of Bibilo Hill, 
was actually 2,500 yards farther on. 

As the regiments advanced, bulldozers 
of the 65th Engineer Battalion at- 
tempted to build jeep trails behind 
them. But the rain fell regularly and 
the trails became morasses so deep that 
even the bulldozers foundered. General 
Collins ordered the trail building 
stopped in mid-August. Now supplies 
were carried by hand and on men's 
backs to the front, and when these meth- 
ods failed to provide enough food and 
ammunition the regiments were sup- 
plied from the air. 



37TH Division Troops Carrying Weapons and Ammunition forward, 5 August 1943. 

The 1st Battalion, 27th Infantry, 
trekked north on the Munda-Bairoko 
trail and made contact on g August with 
Liversedge. The 2d and 3d Battalions, 
27th Infantry, after some sharp fighting 
took Zieta on 15 August, then pushed 
northwest to Piru Plantation. The plan- 
tation lay about seven and one-half air- 
line miles northwest of Bibilo Hill, but 
the regiment's advance on the ground 
required a 22-mile march. The 161st 
Infantry, following the 1st Battalion, 
27 th, moved up the trail and after mid- 
August began patrolling to the west 
short of Bairoko Harbour. On 25 Aug- 
ust, after Griffith had reported two 
nights of busy enemy barge activity in 

and out of Bairoko Harbour, the Amer- 
icans bloodlessly occupied its shores. 

The Japanese Evacuation 

But the main body of Japanese sur- 
vivors had slipped out of Zieta and Bai- 
roko. Traveling light, they had evaded 
the slower-moving, more heavily encum- 
bered Americans. On 5 August General 
Sasaki had decided that he could de- 
fend New Georgia no longer. He there- 
fore sent the 13th Infantry and most of 
the Bairoko-based special naval landing 
force units to Kolombangara, the 229th 
Infantry, the 3d Battalion, 230th In- 
fantry, and the 3d Battalion, 23d In- 
fantry, to Baanga, a long narrow island 



which lay across Lulu Channel from 
Zieta. These units, plus two 120-mm. 
naval guns, were ordered to defend 
Baanga, and the naval guns were to shell 
Munda airfield. 

Sasaki's headquarters, having moved 
out of Munda, was established on Ba- 
anga until 7 August, and the next day 
he moved to Kolombangara. 


The islet of Baanga, 6,500 yards long 
and some 4,000 yards west of Kindu 
Point on New Georgia, was captured to 
extend Allied control over Diamond 
Narrows and to stop the shelling of 
Munda by the two 120-mm. guns nick- 
named "Pistol Pete" by the American 

Seizure of Baanga was entrusted to 
the 43d Division, briefly commanded by 
General Barker after 10 August, when 
General Hodge returned to the Ameri- 
cal Division. 8 Patrolling started on 11 
August, but the Japanese on Baanga 
fought back hard, and the 169th Infan- 
try, which Barker initially assigned to 
Baanga, gave a "shaky performance." 10 
The 17 2d Infantry (less one battalion) 
joined in, and the southern part of the 
island was secured by 21 August. The 
43d Division lost 52 men killed, 110 
wounded, and 486 nonbattle casualties 
in this operation. 

The Japanese, meanwhile, had de- 
cided to get off Baanga. General Sasaki 
had evolved a plan to use the 13th In- 
fantry, then on Kolombangara, to attack 

9 Barker was replaced several days later, on orders 
from the War Department, by General Wing, who 
was senior to him. 

10 See Ltr, Griswold to Harmon, 24 Aug 43, 
quoted in SOPACBACOM, History of the New 
Georgia Campaign, Vol. I, Ch. VII, p. 7, OCMH. 

New Georgia, and dispatched his naval 
liaison officer to 8th Fleet headquarters 
to arrange for air and fleet support. But 
he was to get none. Moreover, no more 
ground reinforcements were to be sent 
to New Georgia. The last attempted 
shipment consisted of two mixed battal- 
ions from the 6th and 38th Divisions, to 
be carried to Kolombangara on destroy- 
ers. 11 But Comdr. Frederick Moosbrug- 
ger with six destroyers surprised the Jap- 
anese force in Vella Gulf on the night 
of 6-7 August and quickly sank three 
Japanese destroyers. The fourth enemy 
ship, which carried no troops, escaped. 
Moosbrugger's force got off virtually 
scot free, while the Japanese lost over 
fifteen hundred soldiers and sailors as 
well as the ships. About three hundred 
survivors reached Vella Lavella. 12 When 
Sasaki's request reached the 8th Fleet, 
Vice Adm. Gunichi Mikawa, basing his 
decision on instructions from Imperial 
General Headquarters, ordered Sasaki to 
cancel his plan for attacking New Geor- 
gia and to concentrate the Baanga troops 
on Arundel to forestall further Allied 
advances. So the Japanese left on barges 
for Arundel, completing the movement 
by 22 August. 

Vella Lavella: The Bypass 

Meanwhile an Allied force had made 
a landing at Barakoma on Vella Lavella, 
which lay about thirty-five nautical miles 
northwest of Munda. This landing rep- 
resented a major and completely success- 
ful departure from the original Toe- 
nails plan. The plan had called for the 
attack against Munda to be followed by 

11 Each battalion consisted of four rifle companies, 
a machine gun platoon, and a small artillery unit. 

12 For a complete account see Morison, Breaking 
the Bismarcks Barrier, pp. 212—22. 



the seizure of Vila airfield on Kolom- 
bangara, but the Japanese were now 
correctly believed to be established on 
Kolombangara in considerable strength. 
Some estimates placed the enemy garri- 
son at ten thousand, a little under the 
actual total. And Admiral Halsey did not 
want a repetition of the Munda cam- 
paign. As he later put it, "The undue 
length of the Munda operation and our 
heavy casualties made me wary of an- 
other slugging match, but I didn't know 
how to avoid it." 13 

There was a way to avoid a slugging 
match, and that way was to bypass Kol- 
ombangara completely and land instead 
on Vella Lavella. The advantages were 
obvious: the airfield at Vila was poorly 
drained and thus no good while Vella 
Lavella looked more promising. Also, 
Vella Lavella was correctly reported to 
contain few Japanese. Vella Lavella, 
northwesternmost island in the New 
Georgia group, lay less than a hundred 
miles from the Japanese bases in the 
Shortlands and southern Bougainville, 
but a landing there could be protected 
by American fighter planes based at 
Munda and Segi Point. 

The technique of bypassing, which 
General MacArthur has characterized as 
"as old as warfare itself," was well under- 
stood in the U.S. Army and Navy long 
before Vella Lavella, but successful by- 
passing requires a preponderance of 
strength that Allied forces had not hith- 
erto possessed. 14 The first instance of an 

u Halsey and Bryan, Admiral Halsey's Story, p. 

"Ltr, MacArthur to Smith, Chief of Mil Hist, 5 
Mar 53, no sub, OCMH. 

In ground operations field orders usually specify 
that isolated pockets of resistance are to be by- 
passed, contained, and reduced later, so that the 
advance will not be held up. 

amphibious bypass in the Pacific oc- 
curred in May 1943 when the Allied cap- 
ture of Attu caused the Japanese to evac- 
uate Kiska. 

When members of Halsey's staff pro- 
posed that South Pacific forces bypass 
Kolombangara and jump to Vella La- 
vella of the more euphonious name, the 
admiral was enthusiastic. 15 On 11 July 
he radioed the proposal to Admirals 
Turner and Fitch. "Our natural route of 
approach from Munda to Bougainville," 
he asserted, "lies south of Gizo and Vella 
Lavella Islands." He asked them to con- 
sider whether it would be practicable 
to emplace artillery in the Munda- 
Enogai and Arundel areas to interdict 
Vila; cut the supply lines to Vila by 
artillery and surface craft, particularly 
PT boats; "by-pass Vila area and allow 
it to die on the vine"; and seize Vella 
Lavella and build a fighter field there. 
The decision on this plan would de- 
pend on the possibility of building a 
fighter field on Vella Lavella to give close 
fighter support for the invasion of Bou- 
gainville. 10 Both Turner and Fitch liked 
the idea. 


Reconnaissance was necessary first. 

" Ltrs, Adm Halsey to Maj Gen Orlando Ward, 
Chief of Mil Hist, 27 May and 27 Aug 52, no subs, 
OCMH. The second letter contains as inclosures 
letters from Admiral Robert B. Carney, formerly 
Chief of Staff, South Pacific Area, and Lt. Gen. 
William E. Riley, USMC (Ret.), formerly Halsey's 
war plans officer, to Admiral Halsey. Other former 
staff officers credit Wilkinson and Comdr. William 
F. Riggs, Jr., with the idea. See Ltr, Vice Adm D. B. 
Duncan to Gen Smith, Chief of Mil Hist, 10 Nov 
53, no sub, OCMH. 

,a Rad, COMSOPAC to CTF 31 and COMAIR- 
SOPAC, n Jul 43, in GHQ SWPA G-3 Jnl, ig Jul 



Allied knowledge of Vella Lavella was 
limited. Coastwatchers, plantation man- 
agers, and such members of the clergy 
as the Rev. A. W. E. Silvester, the New 
Zealand Methodist missionary bishop 
whose see included New Georgia and 
Vella Lavella, provided some informa- 
tion but not enough to form the basis 
for the selection of an airfield site or an 
invasion beach. Col. Frank L. Beadle, 
Harmon's engineer, therefore took com- 
mand of a reconnaissance party consist- 
ing of six Army, Navy, and Marine offi- 
cers. Beadle was ordered to concentrate 
his reconnaissance in the coastal plain 
region of Barakoma and Biloa Mission 
on the southeast tip of Vella Lavella be- 
cause it was closest to Munda, the na- 
tives were friendly, coastwatcher Lt. 
Henry Josselyn of the Australian Navy 
and Bishop Silvester were there, and the 
terrain seemed favorable. The Japanese 
had already surveyed the Barakoma area 
for a fighter strip. 

Beadle's party boarded a torpedo boat 
at Rendova on the night of 21-32 July 
and slipped through the darkness to 
land at Barakoma. Silvester, Josselyn, 
and two natives were on hand to meet 
the American officers. For six days 
Beadle's party, the bishop, the coast- 
watcher, and several natives explored the 
southeast part of the island, and ventured 
up the west coast to Nyanga Plantation, 
about twelve crow's-flight miles north- 
west of Barakoma. Returning to Ren- 
dova on 28 July, Beadle reported that 
the vicinity of Barakoma met all require- 
ments, and that there were no Japanese 
on the southeast coast of the island. 
Beadle recommended that the landing 
be made on beaches extending some 7 50 
yards south from the mouth of the Bara- 

koma River, and suggested that an ad- 
vance detachment be sent to Barakoma 
to mark beaches. These recommenda- 
tions were accepted. 

Admiral Wilkinson, Turner's succes- 
sor, chose an advance party of fourteen 
officers and enlisted men from the var- 
ious units in the Vella Lavella invasion 
force and placed it under Capt. G. C. 
Kriner, USN, who was to command the 
Vella Lavella naval base. This group 
proceeded from Guadalcanal to Ren- 
dova, then prepared to change to PT 
boats for the run through the night of 
12-13 August toward Barakoma. The 
work of the advance party was of a 
highly secret nature. If the Japanese be- 
came aware of its presence, they could 
kill or capture the men and certainly 
would deduce that an Allied invasion 
was imminent. 

On 11 August coastwatcher Josselyn 
radioed Guadalcanal to report the pres- 
ence of forty Japanese. (Japanese sur- 
vivors of the Battle of Vella Gulf, 6-7 
August, had landed on Vella Lavella.) 
His message indicated that pro-Allied 
natives had taken them prisoner. From 
Koli Point General Harmon radioed 
General Griswold to ask for more men 
to accompany the advance party and take 
the prisoners into custody. Accordingly 
one officer and twenty-five enlisted men 
from E and G Companies, 103d In- 
fantry, were detailed to go along. 

The whole party left Rendova at 1730 
on 12 August. En route Japanese planes 
bombed and strafed the four torpedo 
boats for two hours. One was hit and 
four men were wounded but the other 
three made it safely. During the hours 
of darkness they hove to off Barakoma. 
Rubber boats had been provided to get 



the party from the torpedo boats to 
shore, but no one was able to inflate the 
rubber boats from the carbon dioxide 
containers that were provided. So natives 
paddled out in canoes and took the 
Americans ashore. 

Meanwhile Josselyn had radioed Wil- 
kinson again to the effect that there were 
140 Japanese in the area; 40 at Biloa and 
100 about five miles north of Barakoma. 
They were, he declared, under surveil- 
lance but were not prisoners. Once 
ashore Captain Kriner discovered there 
were many starving, ragged, poorly 
armed stragglers but no prisoners. He re- 
quested reinforcements, and in the early 
morning hours of 14 August seventy-two 
officers and enlisted men of F Company, 
103d Infantry, sailed for Barakoma in 
four torpedo boats. This time the rubber 
boats inflated properly and the men 
paddled ashore from three hundred 
yards off the beach. 

The advance party with the secret 
mission of marking beaches and the com- 
bat party with the prisoner-catching mis- 
sion set about their respective jobs. 
Beach marking proceeded in a satis- 
factory manner although the infantry- 
men in that party were not completely 
happy about the presence of the 103d 
troops. They felt that the two missions 
were mutually exclusive and that the 
prisoner-catching mission destroyed all 
hope of secrecy. Only seven Japanese 
were captured. 17 F Company, 103d, held 
the beachhead at Barakoma against the 
arrival of the main invasion force. 


Once Colonel Beadle had made his 
recommendations the various South 
Pacific headquarters began laying their 
plans. This task was fairly simple, for 
Admiral Halsey and his subordinates 
were now old hands at planning in- 
vasions. Actual launching of the in- 
vasion would have to await the capture 
and development of Munda airfield. 

It was on 1 1 August that Halsey issued 
his orders. He organized his forces much 
as he had for the invasion of New Geor- 

gia. (Chart 10) The Northern Force 

" SOPACBACOM's History of the New Georgia 
Campaign, Vol. I, Ch. VIII, pp. 20-21, OCMH, con- 
tains some fanciful data concerning the sinking of 
one PT boat and the killing of about fifty Japanese. 

(Task Force 31) under Admiral Wilkin- 
son was to capture Vella Lavella, build 
an airfield, and establish a small naval 
base. Griswold's New Georgia Occupa- 
tion Force would meanwhile move into 
position on Arundel and shell Vila air- 
field on bypassed Kolombangara. New 
Georgia-based planes would cover and 
support the invasion. South Pacific Air- 
craft (Task Force 33) was to provide air 
support by striking at the Shortlands- 
Bougainville fields. As strikes against 
these areas were being carried out 
regularly, the intensified air operations 
would not necessarily alert the enemy. 
Three naval task forces of aircraft car- 
riers, battleships, cruisers, and destroy- 
ers, and the submarines of Task Force 
72, would be in position to protect and 
support Wilkinson. On Wilkinson's rec- 
ommendation, Halsey set 15 August as D 
Day. 18 

Admiral Wilkinson also issued his 
orders on 1 1 August. The Northern 
Force was organized into three invasion 
echelons (the main body and the second 

I9 COMSOPAC Opn Plan 14A_ 4 3, 11 Aug 43, in 
Off of Naval Reds and Library. 


Chart 10 — South Pacific Organization for Vella Lavella Invasion 


TF 31 


TF 33 

TF 37 

TF 38 

CV. BB, 

TF 39 

TF 72 



and third echelons) and the motor tor- 
pedo boat flotillas. | (Chart n) | Under 
Wilkinson's direct command, the main 
body consisted of three transport groups, 
the destroyer screen, and the northern 
landing force. Each transport group, 
screened by destroyers, was to move in- 
dependently from Guadalcanal to Vella 
Lavella; departure from Guadalcanal 
would be so timed that each group would 
arrive off Barakoma just before it was 
scheduled to begin unloading. Three 
slow LST's, each towing an LCM, would 
leave at 0300, 14 August, six LCI's at 
0800, and seven fast APD's at 1600. The 
motor torpedo boat flotilla would cover 
the movement of the main body on D 
minus 1 by patrolling the waters east 
and west of Rendova, but would retire 
to Rendova early on D Day to be out of 
the way. Preliminary naval bombard- 
ment would in all probability not be 

necessary, but Wilkinson told off two 
destroyers to be prepared to support the 
landing if need be. Two fighter-director 
groups were put aboard two destroyers. 
Once unloaded, each transport group 
would steam for Guadalcanal. The sec- 
ond echelon, composed of three LST's 
and three of the destroyers that would 
escort the main body, was to arrive at 
Barakoma on D plus 2, beach overnight, 
and return to Guadalcanal. The third 
echelon consisted of three destroyers and 
three LST's from the main body. Wil- 
kinson ordered it to arrive on D plus 5, 
beach throughout the night, and depart 
for Guadalcanal the next morning. 19 

The northern landing force, 5,888 
men in all, consisted of the 35th Regi- 
mental Combat Team of the 25th Di- 

CTF 31 Opn Plan Aia-43, 11 Aug 43, in Off of 
Naval Reds and Library. 

Chart 11— Organization of Northern Force [TF 31], Vella Lavella 

Northern Force 
Adm Wilkinson 

TC 31.5 
Main Body 

TU 31.5.1 
Adv Transport 

7 APO'i; 6 DD'i 

TU 31.5.2 
2d Transport 

12 LCI's, 4 OD's 

TG 31 .2 
12 DD's i 

Gen McClure 


TG 31.6 
2d Ecfielon 
Cap! Cooke 

TG 31.7 
3d Echelon 
Cap) Carter 

TU 31.5.3 
3d Transport Group 
3 LST's, 2 DD's, 
2 SO 


3 LST's; 3 DD's 
1 SC 

Naval base 

Copt Kriner 

Hq Det 58th NCB (-) 

35th RCT (- det,) Nov base uniti 

35th Inl, Col Everett E. Brown 
64th FA Bn 
C Co, 65th L™ . 
Collectine Co B, 25th Med Bn 


Df, 35th RCT 

155-mm. Gun Gp, 
4th Def Bn 
Blry (90 -mm.)^ 

4th Del Bn 
Del 58th NC8 

TG 31. B 
MTfl Flotilla 

TU 31.8.1 
New Georgia MTB 
1 8 MTB's 

Land ins 

TU 31,8.2 
8 MTB's ' 

Del 35th RCT 

Tk PI, 4th Mar Del Bn 

Del 58th NCB 



vision; the 4th Marine Defense Bat- 
talion; the 25 th Cavalry Reconnaissance 
Troop; the 58th Naval Construction 
Battalion; and a naval base group. 20 
Command of the landing force was en- 
trusted to Brig. Gen. Robert B. Mc- 
Clure, assistant commander of the 25th 
Division, who as a colonel had com- 
manded the 35th Infantry during the 
Guadalcanal Campaign. General Mc- 
Clure would be under Wilkinson's con- 
trol until he was well established ashore. 
He would then come under General 

The Japanese on Vella Lavella (no 
garrison at all but only a group of 
stragglers) were estimated to total about 
250, with 100 more on nearby Ganongga 
and 250 at Gizo. Wilkinson warned that 
enemy air strength in southern Bougain- 
ville, less than a hundred miles away, 
and at Rabaul was considerable, and that 
naval surface forces were based at both 

To carry off such a stroke almost 
literally under the enemy's aircraft 
would require, besides fighter cover, 
considerable speed in unloading. Wil- 
kinson planned to unload the main body 
in twelve hours. Troops debarking from 
APD's were to go ashore in LCVP's, 
forty to a boat. At the beach ten of each 
forty would unload the boat while the 

10 The 35th Regimental Combat Team consisted 
of the 35th Infantry; the 64th Field Artillery Bat- 
talion (105-mm. howitzer); C Company, 65th En- 
gineer Battalion; Collecting Company B, 25th Med- 
ical Battalion, and detachments from other di- 
visional services. Harmon, who had promised the 
35th Infantry for New Georgia on 1 August, later 
considered using the 145th Infantry, but concluded 
that it could not be pulled out of New Georgia and 
brought back to Guadalcanal in time. Rads, Harmon 
to Griswold, 1 Aug and 6 Aug 43, in XIV Corps 
G-3 Jnl. 

thirty pushed inland. Once emptied, 
LCVP's were to return to their mother 
ships for the rest of the men and sup- 
plies. Sixty minutes were allotted for un- 
loading the APD's and clearing the 
beach. The LCI's would then come in 
to the beach and drop their ramps. Pas- 
senger troops would debark via both 
ramps, ground their equipment, then re- 
board by the starboard ramps, pick up 
gear, and go ashore down the port ramps. 
One hour was allotted for the LCI's. 
Then the LST's, bearing artillery, 
trucks, and bulldozers, would ground. 
Trucks were to be loaded in advance to 
help insure the prompt unloading of the 

The 35 th Infantry, commanded by 
Col. Everett E. Brown, had been mak- 
ing ready for several days. It had 
been alerted for movement to Munda 
on 1 August, and on 9 August had re- 
ceived orders from Harmon's headquart- 
ers to prepare for an invasion. The 1st 
and 2d Battalions on Guadalcanal and 
the 3d Battalion and the 64th Field Artil- 
lery Battalion in the Russells then began 
rehearsing landings. In the week preced- 
ing the invasion South Pacific Aircraft 
struck regularly at Kolombangara, Buin, 
Kahili, and Rekata Bay. 

By 14 August the landing force and 
its supplies were stowed aboard ship, 
and all transport groups of the main 
body shoved off for Barakoma on sched- 
ule. Once on board, the men were in- 
formed of their destination. Japanese 
planes were reported over Guadalcanal, 
the Russells, and New Georgia, but Wil- 
kinson's ships had an uneventful voy- 
age up the Slot and through Blanche 
Channel and Gizo Strait. The sea was 
calm, and a bright moon shone in the 



clear night sky. Northwest of Rendova 
the LCrs overhauled the LST's while 
the APD's passed both slower groups. 

Seizure of Barakoma 

As first light gave way to daylight in 
the morning of 15 August, the APD's 
carrying the 1st and 2d Battalions of the 
35th Infantry arrived off Barakoma and 
hove to. General Mulcahy's combat air 
patrol from Munda and Segi Point 
turned up on schedule at 0605. With 
part of the 103d Infantry and the secret 
advance party already on shore and in 
possession of the landing beach, there 
was no need for support bombardment. 
The APD's swung landing craft into the 
water, troops of the 2d Battalion climbed 
down the cargo nets and into the boats, 
and the first wave, with rifle companies 
abreast, departed. The 2d Battalion hit 
the beach at 0624 and at once pushed 
south toward the coconut plantation 
around Biloa Mission, about four thou- 
sand yards from the beach. The 1st Bat- 
talion, having left the APD's at 0615, 
landed with companies abreast at 0630 
and pushed northward across the waist- 
deep Barakoma River. Thus quickly un- 
loaded, the APD's cleared the area and 
with four escorting destroyers proceeded 
toward Guadalcanal. 

The twelve LCI's arrived on schedule 
and sailed in to the beach, but quickly 
found there was room for but eight at 
one time. Coral reefs a few yards from 
shore rendered the northern portion of 
the beach unusable. The remaining four 
LCI's had to stand by offshore awaiting 
their turn. The 3d Battalion started un- 
loading but had barely gotten started 
when enemy aircraft pounced at the in- 
vasion force. 

This time the Japanese were not 
caught so completely asleep as they had 
been on 30 June. In early August Japa- 
nese radio intelligence reported a good 
deal of Allied radio traffic, and the 
commanders at Rabaul were aware that 
ships were again concentrating around 
Guadalcanal. They concluded that a new 
invasion was impending but failed to 
guess the target. At 0300 of the 15th a 
land-based bomber spotted part of Wil- 
kinson's force off Gatukai, Six dive 
bombers and forty-eight fighters were 
sent out on armed reconnaissance, and 
these found the Americans shortly be- 
fore 0800. Mulcahy's planes and ships' 
antiaircraft guns promptly engaged 
them. The Japanese planes that broke 
through went for the destroyers, which 
steamed on evasive courses and escaped 
harm. The Japanese caused some casual- 
ties ashore by strafing, but did not attack 
the LST's and LCI's. They ebulliently 
reported repulsing fifty Allied planes. 

This attack, together with the limita- 
tions of the beach, delayed the unloading 
of the LCI's, which did not pull out un- 
til 0915. 21 The three LST's then beached 
and began discharging men and heavy 
cargo. Unloading continued all day. 

The Japanese struck again at 1227; 
eleven bombers and forty-eight fighters 
came down from the north. Some at- 
tacked the LST's but these "Large, Slow 
Targets" had mounted extra antiaircraft 
guns and brought down several Japa- 
nese planes. 

At 1724, some thirty-six minutes be- 
fore the LST's departed, the enemy came 
again. Forty-five fighters and eight bomb- 

21 The 35th Infantry later reported the existence 
of a longer beach eight hundred yards north of 



ers attacked without success. The Japa- 
nese pilots who flew against the Northern 
Force on that August day showed a 
talent for making unwarranted claims. 
A postwar account soberly admits the 
loss of seventeen planes, but claims the 
sinking of four large transports, one 
cruiser, and one destroyer. It states that 
twenty-nine Allied planes were shot 
down and that four large transports were 
damaged. 22 The ships retiring from Vella 
Lavella were harried from the air almost 
all the way, fortunately without suffering 

D Day, a resounding success, had pro- 
ceeded with the efficiency that charac- 
terized all Admiral Wilkinson's opera- 
tions. Landed were 4,600 troops (700 of 
whom were naval personnel) and 2,300 
tons of gear including eight go-mm. anti- 
aircraft guns, fifteen days' supplies, and 
three units of fire for all weapons except 
antiaircraft guns, for which one unit was 
landed. The 35th Regimental Combat 
Team established a perimeter defense. 
Field artillery was in position by 1700, 
and by 1530 the 4th Marine Defense Bat- 
talion had sixteen .50-caliber, eight 20- 
mm., and eight 40-mm. antiaircraft guns 
and two searchlights in place. The guns 
engaged the last flight of enemy planes. 
There were some problems, of course. 
The LST's had been unloaded slowly, 
but supplies came ashore faster than the 
shore party could clear them off the 
beach. Boxes of equipment, ammuni- 
tion, and rations were scattered about. 
The troops had brought their barracks 
bags and these lay rain-soaked in the 
mud. Unused field stoves stood in the 
way for several days. 

The bypass to Vella Lavella was easier 
and cheaper than an assault on Kolom- 
bangara. Twelve men were killed and 
forty wounded by air bombing and straf- 
ing, but D Day saw no fighting on the 

There was never any real ground com- 
bat on Vella Lavella, because Japanese 
stragglers were mainly interested in es- 
cape rather than fighting. When it be- 
came known that Wilkinson was landing 
on Vella Lavella officers of the 8th Fleet 
and the rjth Army went into conference. 
They estimated, with an accuracy un- 
usual for Japanese Intelligence, that the 
landing force was about a brigade in 
strength. With blithe sanguinity some- 
one proposed sending a battalion to 
effect a counterlanding. General Ima- 
mura's headquarters took a calmer view 
and pointed out that sending one bat- 
talion against a brigade would be "pour- 
ing water on a hot stone." 23 The 8th 
Area Army stated that two brigades 
would be needed to achieve success, but 
that not enough transports were avail- 
able. In view of the general Japanese 
strategy of slow retreat in the central 
Solomons in order to build up the de- 
fenses of Bougainville and hold Rabaul, 
it was decided to send two rifle com- 
panies and one naval platoon to Horaniu 
at Kokolope Bay on the northeast corner 
of Vella Lavella to establish a barge stag- 
ing base between the Shortlands and 

The real struggle for Vella Lavella 
took place in the air and on the sea. 
Japanese naval aircraft made a resolute 
effort to destroy the American ships 
bearing supplies and equipment to Vella. 

"Southeast Area Naval Operations, II, Japanese 
Monogr No. 49 (OCMH), 47-48. 

Ibid., p. 48. 



Fighters and bombers delivered daylight 
attacks and seaplanes delivered a series 
of nocturnal harassing attacks that were 
all too familiar to Allied troops who 
served in the Solomons in 1942, 1943, 
and early 1944. 

The combat air patrol from Mulcahy's 
command made valiant efforts to keep 
the Japanese away during daylight, but 
as radar-equipped night-fighters did not 
reach the New Georgia area until late 
September shore- and ship-based antiair- 
craft provided the defense at night. 24 For 
daylight cover Mulcahy had planned to 
maintain a 32-plane umbrella over Bara- 
koma, but the limited operational fa- 
cilities at Munda made this impossible 
at first. On 17 August only eight fighters 
could be sent up at once to guard Bara- 
koma. To add to the difficulties, the 
weather over New Georgia was bad for 
a week, the fighter-director teams on 
Vella Lavella were new to their task, and 
one of the 4th Defense Battalion's radars 
was hit by a bomb on 17 August. 

The Northern Force's second echelon, 
under Capt. William R. Cooke, having 
departed Guadalcanal on 16 August, 
beached at Barakoma at 1626 the next 
day. The fighter cover soon left for 
Munda and at 1850, and again at 1910, 
Japanese planes came over to bomb and 
strafe. General McClure ordered Cooke 
not to stay beached overnight but to put 
to sea. Escorted by three destroyers, the 
LST's went down Gizo Strait where they 
received an air attack of two hours and 
seventeen minutes' duration. The con- 
voy sailed toward Rendova until 0143, 
then reversed and headed for Barakoma. 

"Robert Sherrod, History of Marine Corps Avi- 
ation in World War II (Washington: Combat 
Forces Press, 1952), p. 163. 

One LST caught fire, probably as a re- 
sult of a gasoline vapor explosion, and 
was abandoned without loss of life. The 
other two reached Barakoma, suffered 
another air attack, unloaded, and re- 
turned safely to Guadalcanal. 

Capt. Grayson B. Carter led the third 
echelon to Vella Lavella. It was attacked 
by enemy planes in Gizo Strait at 0540 
on 21 August; one destroyer was slightly 
damaged and two men were killed. 
Planes struck off and on all day at the 
beached LST's, but men of the 58th 
Naval Construction Battalion showed 
such zeal in unloading that the LST's 
were emptied by 1600. The next con- 
voys, on 26 and 31 August, had less ex- 
citing trips. The weather had cleared 
and the air cover was more effective. 
During the first month they were there, 
the Americans on Vella Lavella re- 
ceived 108 enemy air attacks, but none 
caused much damage. In the period be- 
tween 15 August and 3 September, the 
day on which Wilkinson relinquished 
control of the forces ashore, Task Force 
31 carried 6,505 men and 8,626 tons of 
supplies and vehicles to Vella Lavella. 

During that period General McClure's 
troops had strengthened the defenses 
of Barakoma, established outposts and 
radar stations, and patrolled northward 
on both coasts. On 28 August a thirty- 
man patrol from the 25th Cavalry Re- 
connaissance Troop that had accom- 
panied radar specialists of the 4th Marine 
Defense Battalion in search of a new 
radar site reported considerable enemy 
activity at Kokolope Bay. 

Capture of Horaniu 

Having decided to establish the barge 
base at Horaniu, the Japanese sent the 




Warship Firing at Japanese Destroyers near f/ie coasJ o/ F<?Wa Lavella, early 
morning, 18 August. 

two Army companies and the naval pla- 
toon, 390 men in all, out of Erventa on 
17 August. Four torpedo boats, 13 troop- 
carrying daihatsu barges, 25 z armored 
barges, 2 submarine chasers, 1 armored 
boat, 4 destroyers, and 1 naval air group 
from the Shortlands were involved. The 
destroyers were intercepted north of 
Vella Lavella in the early morning hours 
of 18 August. The daihatsus dispersed. 
The Americans sank the 2 submarine 

25 Daihatsu is an abbreviation for Ogata Hatsu- 
dokitei which means a large landing barge. The 
daihatsu was 41—44 feet long; it could carry 100- 
120 men for short distances, 40-50 on long trips. 
The sides were usually armored, and it carried 
machine guns. 

chasers, ? motor torpedo boats, and 1 
barge. The Japanese destroyers, two of 
which received light damage, broke off 
the action and headed for Rabaul. 26 
Harried by Allied planes, the daihatsus 
hunted for and found Horaniu, and the 
troops were ashore by nightfall on 19 
August. About the same time, General 
Sasaki took alarm at the seizure of Bara- 
koma and sent the 2d Battalion, 45th 
Infantry, and one battery of the 6th 
Field Artillery Regiment from Kolom- 
bangara to defend Gizo. 

When General McClure received the 

M For a complete account see Morison, Breaking 

the Bismarcks Barrier, pp. 234-37. 



report from the reconnaissance troop 
patrol on 28 August, he ordered Maj. 
Delbert Munson's 1st Battalion to ad- 
vance up the east coast and take the 
shore of Kokolope Bay for a radar site. 
To take the 1st Battalion's place in the 
perimeter defense, McClure asked Gris- 
wold for a battalion from New Georgia. 
The 1st Battalion, 145th Infantry, was 

On the morning of 30 August Major 
Munson dispatched A Company up the 
east coast ahead of his battalion, and 
next day, after the arrival of the 1st Bat- 
talion, 145th Infantry, the main body of 
the 1st Battalion, 35th, started north. 
Josselyn and Bishop Silvester had pro- 
vided native guides and the bishop gave 
Munson a letter instructing the natives 
to help the American soldiers haul sup- 
plies. C Company, 65th Engineer Bat- 
talion, was to build a supply road behind 

By afternoon of 1 September A Com- 
pany had reached the vicinity of Orete 
Cove, about fourteen miles northeast of 
Barakoma. The main body of the bat- 
talion was at Narowai, a village about 
seven thousand yards southwest of Orete 
Cove. The coastal track, which had been 
fairly good at first, narrowed to a trail 
that required the battalion to march in 
single file. Inland were the jungled 
mountains of the interior. Supply by 
hand-carry was impossible, and Mc- 
Clure and Colonel Brown, who had been 
informed that higher headquarters ex- 
pected the Japanese to evacuate, decided 
to use landing craft to take supplies to 
Munson. On 2 September supplies ar- 
rived at Orete Cove along with seven- 
teen Fiji scouts under Tripp, who had 
recently been promoted to major. 

From that day until 14 September 
Munson's battalion, supported by the 3d 
Battalion, 35th, and C Battery, 64th Field 
Artillery Battalion, moved forward in a 
series of patrol actions and skirmishes. 

Horaniu fell on 14 September. The 
Japanese did not seriously contest the 
advance. Instead they withdrew steadily, 
then moved overland to the northwest 
corner of the island. 

Up to now troops of the United States 
had borne the brunt of ground combat 
in the Solomons, but Admiral Halsey had 
decided to give the 3d New Zealand Di- 
vision a chance to show its mettle. He 
had earlier moved the division from New 
Caledonia to Guadalcanal. On 18 Sep- 
tember Maj. Gen. H. E. Barrowclough, 
general officer commanding the division, 
took over command of Vella Lavella 
from General McClure. On the same day 
the 14th New Zealand Brigade Group 
under Brigadier Leslie Potter landed 
and began the task of pursuing the re- 
treating enemy. 27 Battalion combat teams 
advanced up the east and west coasts, 
moving by land and by water in an at- 
tempt to pocket the enemy. But the Japa- 
nese eluded them and got safely oflf the 

The Seabees had gone to work on the 
airfield at once. As at Munda, good coral 
was abundant. By the end of August they 
had surveyed and cleared a strip four 
thousand feet long by two hundred feet 
wide. They then began work on a control 
tower, operations shack, and fuel tanks. 

" A brigade group was similar in strength and 
composition to a U.S. regimental combat team. For 
details of Potter's operations see Oliver A. Gillespie, 
The Pacific, "The Official History of New Zealand 
in the Second World War, 1939-1945" (Wellington, 
New Zealand, 1952), pp. 125-42. 



The first plane to use the field landed 
on 24 September, and within two months 
after the invasion the field could accom- 
modate almost a hundred aircraft. 

The decision to bypass Kolombangara 
yielded this airfield in return for a low 
casualty rate. Of the Americans in the 
northern landing force, 19 men were 
killed by bombs, 7 died from enemy gun- 
fire, and 108 were wounded. Thirty-two 
New Zealanders died, and 32 were 

Final Operations 


About the time that Vella Lavella was 
being secured, General Griswold's forces 
on New Georgia were carrying out their 
part of Admiral Halsey's plan by seizing 
Arundel and by shelling Kolombangara 
to seal it off. The attack on Arundel, 
which is separated from the west coast of 
New Georgia by Hathorn Sound and 
Diamond Narrows, proved again that it 
was all too easy to underestimate the 
Japanese capacity for resolute defense. 
The i72d Infantry invaded it on 27 
August, but the Japanese fought so 
fiercely that the 27th Infantry, two bat- 
talions of the i6gth Infantry, one com- 
pany of the 103d Infantry, B Company 
of the 82d Chemical Battalion (4.2-inch 
mortars, in their South Pacific debut), 
the 43d Reconnaissance Troop, and six 
Marine tanks had to be committed to 
keep the offensive going. 

Resistance proved more intense than 
expected in part because the indefati- 
gable Sasaki had not yet abandoned his 
hope of launching an offensive that 
would recapture Munda. On 8 Septem- 
ber he sent the 3d Battalion, 13th In- 
fantry, from Kolombangara to strengthen 

his forces on Arundel, and five days later, 
when Allied air and naval forces had 
practically cut the supply lines between 
Bougainville and Kolombangara and his 
troops faced starvation, he decided to at- 
tack Munda or Bairoko via Arundel and 
seize the Americans' food. He therefore 
dispatched Colonel Tomonari (who was 
slain in the ensuing fight) and the rest of 
the 13 th Infantry to Arundel on 14 Sep- 

Thus the battle for Arundel lasted un- 
til 21 September, and ended then, with 
the Americans in control, only because 
Sasaki ordered all his Arundel troops to 
withdraw' to Kolombangara. 

The Japanese Evacuation 

Sasaki had ordered the evacuation of 
Arundel because Imperial General Head- 
quarters had decided to abandon the 
New Georgia Islands completely. While 
the Americans were seizing Munda air- 
field, the Japanese naval Authorities in 
the Southeast Area realized that their 
hold on the central Solomons was tenu- 
ous. But they resolved to maintain the 
line of communications to Kolomban- 
gara, so that Sasaki's troops could hold 
out as long as possible. If Sasaki could 
not hold out, the next best thing would 
be a slow, fighting withdrawal to buy 
time to build up defenses for a final stand 
on Bougainville. 

Such events in early August as the fall 
of Munda and the Japanese defeat in 
Vella Gulf on 6-7 August precipitated 
another argument between Japanese 
Army and Navy officers over the relative 
strategic merits of New Guinea and the 
Solomons. This argument was resolved 
in Tokyo by the Imperial General Head- 
quarters which decided to give equal 



14TH New Zealand Brigade Group landing on Vella Lavella, 18 September 1943. 

priority to both areas. Tokyo sent orders 
to Rabaul on 13 August directing that 
the central Solomons hold out while Bou- 
gainville was strengthened, and that the 
central Solomons were to be abandoned 
in late September and early October. The 
decision to abandon New Georgia was 
not made known at once to General 

Sasaki, with about twelve thousand 
men concentrated on Kolombangara, 
prepared elaborate defenses along the 
southern beaches and, as shown above, 
prepared plans for counterattacks. 
Finally on 15 September, after Sasaki 
had sent the 13th Infantry to Arundel, 
an 8th Fleet staff officer passed the word 
to get his troops out. 

Southeastern Fleet, 8th Fleet, and 

Sasaki's headquarters prepared the plans 
for the evacuation. A total of 12,435 men 
were to be moved. Eighteen torpedo 
boats, thirty-eight large landing craft, 
and seventy or eighty Army barges 
(daihatsus) were to be used. 28 Destroyers 
were to screen the movement, aircraft 
would cover it, and cruisers at Rabaul 
would stand by in support. 

The decision to use the daihatsus was 
logical, considering the destroyer losses 
in Vella Gulf and the success the noc- 
turnal daihatsus had enjoyed. American 

" 17th Army Operations, II, Japanese Monogr No. 
40 (OCMH), 54, says 138 "large motor boats" were 
to be used; Southeast Area Naval Operations, II, 
Japanese Monogr No. 49 (OCMH), 52, lists 18 tor- 
pedo boats, 38 large landing barges, and about 70 
Army craft. 



PT boat squadrons, four in all, had been 
operating nightly against the enemy 
barges in New Georgian waters since late 
July, and had sunk several, but only a 
small percentage of the total. Destroyers 
and planes had also operated against 
them without complete success. The 
Japanese put heavier armor and weapons 
on their barges for defense against tor- 
pedo boats, which in turn replaced their 
torpedoes— useless against the shallow- 
draft barges— with 37-mm. antitank and 
40-mm. antiaircraft guns. The barges 
were too evasive to be suitable targets 
for the destroyers' 5-inch guns. Planes of 
all types, even heavy bombers, hunted 
them at sea, but the barges hid out in the 
daytime in carefully selected staging 
points. Those traveling by day covered 
themselves with palm trees and foliage 
so that from the air they resembled islets. 

Sasaki ordered his troops off Gizo and 
Arundel; those on Arundel completed 
movement to Kolombangara by 21 Sep- 
tember. The seaplane base at Rekata Bay 
on Santa Isabel and the outpost on Ga- 
nongga Island were also abandoned at 
this time. The evacuation of Kolomban- 
gara was carried out on the nights of 
28-29 September, 1-2 October, and 2-3 
October. Admiral Wilkinson had antici- 
pated that the enemy might try to escape 
during this period, the dark of the moon. 
Starting 22 September American cruisers 
and destroyers made nightly reconnais- 
sance of the Slot north of Vella Lavella, 
but when Japanese submarines became 
active the cruisers were withdrawn. The 
destroyers attempted to break up the 
evacuation but failed because enemy 
planes and destroyers interfered. The 
Japanese managed to get some 9,400 men, 
or some 3,600 less than they had evacu- 

ated from Guadalcanal in February, 
safely off the island. Most of them were 
sent to southern Bougainville. Twenty- 
nine landing craft and torpedo boats 
were sunk, one destroyer was damaged, 
and sixty-six men were killed. 29 

The final action in the New Georgia 
area was the Battle of Vella Lavella on 
the night of 6-7 October, when ten Japa- 
nese destroyers and twelve destroyer- 
transports and smaller craft came down 
to Vella Lavella to rescue the six hun- 
dred stranded men there. Facing odds of 
three to one, American destroyers en- 
gaged the Japanese warships northwest 
of Vella Lavella. One Japanese destroyer 
was sunk; one American destroyer was 
badly damaged and sank, and two more 
suffered damage. During the engagement 
the transports slipped in to Marquana 
Bay on northwest Vella Lavella and got 
the troops out safely. 30 The last organized 
bodies of Japanese had left the New 
Georgia area. 

When the 1st Battalion, 27th In- 
fantry, landed at Ringi Cove on southern 
Kolombangara on the morning of 6 
October, it found only forty-nine aban- 
doned artillery pieces and some scattered 
Japanese who had been left behind. 31 
The long campaign— more than four 
months had elapsed since the Marines 
landed at Segi Point— was over. 

29 Southeast Area Naval Operations, II, Japanese 
Monogr No. 49 (OCMH), 54—55. At the time the 
Americans greatly overestimated their success against 
barges. See for example Halsey and Bryan, Admiral 
Halsey's Story, p. 172. 

30 See Morison, Breaking the Bismarcks Barrier, 
pp. 343-52, for a more complete account. 

31 ACofS G-2 XIV Corps to CG XIV Corps, 29 Oct 
43, sub: Photo Int Rpt, Kolombangara, with incls, 
Exhibit 6, List of Guns Abandoned by Enemy in 
the Vila Area. 



Table 2— American Casualties on New Georgia 






Killed in action and died of wounds.. 


















Source: NGOF, Narrative Account of the Campaigns in the New Georgia Group, p. 29. 


New Georgia had been lengthy and 
costly. Planned as a one-division affair, 
it had used up elements of four di- 
visions. It would be months before the 
25th and 43d Divisions were ready to 
fight again. American casualties totaled 
1,094 dead, 3,873 wounded. (Table 2) 
These figures do not tell the complete 
story, for they count only men killed or 
wounded by enemy fire. They do not in- 
clude casualties resulting from disease or 
from combat fatigue or war neuroses. For 
example the i72d Infantry reported 
1,550 men wounded or sick; the 169th 
Infantry, up to 5 August, suffered 958 
nonbattle casualties. The 103d Infantry 
had 364 "shelled-shocked" and 83 non- 
battle casualties. 32 

Japanese casualties are not known, but 
XIV Corps headquarters reported a 
count of enemy dead, exclusive of Vella 
Lavella, of 2,483. 

The Allied soldiers, airmen, marines, 
and sailors who suffered death, wounds, 
or illness, and those who fought in the 
campaign without injury, had served 
their cause well. New Georgia was a suc- 

32 i?2d Inf Rpt of Opns, New Georgia; 169th Inf 
Jnl, 5 Aug 43; 103d Inf Rpt of Opns, New Georgia. 

cess. The bypassing of Kolombangara, 
though overshadowed by later bypasses 
and clouded by the fact that the bypassed 
troops escaped, was a satisfactory demon- 
stration of the technique; the seizure of 
Vella Lavella provided Halsey's forces 
with a good airfield for a much lower 
price in blood than an assault on Kolom- 
bangara. The Allies swiftly built another 
airfield at Ondonga Peninsula on New 
Georgia. This gave them four— Munda, 
Barakoma, Ondonga, and Segi. The first 
three, the most used, brought all Bou- 
gainville within range of Allied fighters. 
When South Pacific forces invaded the 
island, they could pick an undefended 
place and frustrate the Japanese efforts 
to build up Bougainville's defenses and 
delay the Allies in New Georgia. 

The New Georgia operation is also 
significant as a truly joint operation, and 
it clearly illustrates the interdependence 
of air, sea, and ground forces in oceanic 
warfare. Victory was made possible only 
by the close co-ordination of air, sea, and 
ground operations. Air and sea forces 
fought hard and finally successfully to cut 
the enemy lines of communication while 
the ground troops clawed their way for- 
ward to seize objectives intended for use 
by the air and sea forces in the next ad- 
vance. Unity of command, established 



Ma j. Gen. J. Lawton Collins talking to Maj. Charles W. Davis, Commanding 
Officer, 3d Battalion, 2jth Infantry (left), New Georgia, August 1943. 

from the very start, was continued 
throughout with obvious wholehearted- 
ness by all responsible commanders. 

No account of the operation should 
be brought to a close without praising 
the skill, tenacity, and valor of the 
heavily outnumbered Japanese who 

stood off nearly four Allied divisions in 
the course of the action, and then suc- 
cessfully evacuated 9,400 men to fight 
again. The obstinate General Sasaki, who 
disappears from these pages at this point, 
deserved his country's gratitude for his 
gallant and able conduct of the defense. 


The Markham Valley and the 
Huon Peninsula 

While South Pacific troops had been 
so heavily engaged in New Georgia, Gen- 
eral MacArthur's Southwest Pacific forces 
were executing Operation II of the 
Elkton plan— the seizure of the Mark- 
ham Valley and the Huon Peninsula of 
New Guinea— aimed at increasing the 
Southwest Pacific Area's degree of con- 
trol over Vitiaz and Dampier Straits. 
(Map 12) This operation had actually 
started in January 1943 with the Aus- 
tralian defense of Wau in the Bulolo 
Valley, and was furthered by the Aus- 
tralian advance from the Bulolo Valley 
toward Salamaua and the 30 June land- 
ing of the MacKechnie Force at Nassau 


The Allies 

The ground forces in Operation II (or 
Postern) were under command of the 
New Guinea Force. 1 General Blarney ar- 
rived at Port Moresby and assumed com- 

1 The subsection is based on ALF, Rpt on New 
Guinea Opns: 4 Sep 43-26 Apr 44; GHQ SWPA 
Warning Instns 2, 6 May 43, in GHQ SWPA G-3 
Jnl, 6 May 43; GHQ SWPA OI 34, 13 Jun 43, and 
subsequent amendments, in GHQ SWPA G-3 Jnl, 
14 Jun 43; Craven and Cate, The Pacific: Guadal- 
canal to Saipan, pp. 183-86; Kenney, General Ken- 
ney Reports, pp. 273—87; Morison, Breaking the 

mand of the New Guinea Force on 20 
August 1943, and General Herring went 
to Dobodura, where as general officer 
commanding the I Australian Corps he 
exercised control over tactical operations. 
General Blarney was responsible for co- 
ordination of ground, air, and naval 
planning. In the actual conduct of 
ground, air, and naval operations, the 
principle of co-operation rather than 
unity of command appears to have been 

Bismarcks Barrier, Ch. XIV; Memo, ACofS G-3 
GHQ SWPA for CofS GHQ SWPA, 14 Jul 43, no 
sub, in GHQ SWPA G-3 Jnl, 14 Jul 43; Ltr, Adv 
Hq ALF to GHQ SWPA, 16 Jul 43, sub: Opns 
Cartwheel, in GHQ SWPA G-3 Jnl, 16 Jul 43; 
Ltr, Brig Gen Donald Wilson, CofS AAF SWPA, 
to CINCSWPA, 20 Jul 43, sub: Supporting Plan, 
GHQ OI 34, same file; LHQ [ALF] OI 54, 30 Jul 
43, in GHQ SWPA G-3 Jnl, 31 Jul 43; NGF OI gg, 
25 Aug 43, in GHQ SWPA G-3 Jnl, 25 Aug 43; 
Memo, Gen Chamberlin, ACofS G-3 GHQ SWPA, 
for CofS GHQ SWPA, 25 May 43, sub: Control of 
Opns of 2d ESB, in GHQ SWPA G-3 Jnl, 25 May 
43; ANF Opn Plan 5-43, 19 Jul 43, in GHQ SWPA 
G-3 Jnl, 20 Jul 43;' AAF SWPA OI 37, 18 Jun 43, 
in GHQ SWPA G-3 Jnl, 19 Jun 43; Ltr, Comdr 
ANF to CINCSWPA, 16 Aug 43, sub: Air Support 
for Troop and Sup Overwater Movements During 
Postern Opn, in GHQ SWPA G-3 Jnl, 16 Aug 43; 
Memo, Gen Kenney for Gen Chamberlin, 25 Aug 
43, sub: Opn Plan, Adv Ech 15th AF, Postern Opn, 
same file; G-2 Est Postern, 20 Aug 43, in GHQ 
SWPA G-3 Jnl, 20 Aug 43. 



,• % g Mt Gkidstone\ 

•• sit 

t % 






SoKor I 


Gloucester $ 




V i t o Islands 





MAP 12 

Printed by Army Map Service X>- Vlolmes, ]t. 



The operations involved in the seizure 
of the Huon Peninsula and the Mark- 
ham Valley were complex. The South- 
west Pacific lacked enough ships for a 
completely amphibious assault, and had 
too few aircraft for a completely air- 
borne attack; there were enough ground 
troops, but New Guinea terrain pre- 
cluded large-scale overland movements. 
To bring sufficient power to bear General 
MacArthur and his subordinates and 
staff therefore employed all available 
means—amphibious assault, an assault 
by parachute troops, an airlift of an 
entire division, and the shore-to-shore 
operation already executed at Nassau 

MacArthur, in operations instructions 
issued before the invasions of Woodlark, 
Kiriwina, and Nassau Bay, and followed 
by a series of amendments, ordered the 
New Guinea Force to seize the Lae- 
Markham Valley area by co-ordinated 
airborne and overland operations 
through the Markham Valley and am- 
phibious operations (including Nassau 
Bay) along the north coast of New 
Guinea. The Markham Valley operations 
were to be based on Port Moresby; the 
north coast operations on Buna and 
Milne Bay. MacArthur directed the 
seizure of the coastal areas of the Huon 
Gulf, including Salamaua and Finsch- 
hafen, and initially ordered the New 
Guinea Force to be prepared for air- 
borne-overland and shore-to-shore opera- 
tions along the north coast of New 
Guinea as far as Madang on Astrolabe 
Bay. The immediate objectives were Lae 
and the Markham and Rarau Valleys. 

The two river valleys form a tremen- 
dous trough between the Finisterre and 
Kratke Ranges. Starting at the mouth of 

the Markham River at Lae and running 
northwesterly for 380 miles to the mouth 
of the Ramu, the trough varies from 5 
to 25 miles in width. The rivers flow in 
opposite directions from a plain in the 
level uplands of the trough some 80 miles 
northwest of Lae. Both valleys contain 
extensive flats of grass-covered sand and 
gravel, and thus there were many excel- 
lent sites for air bases. Already in exist- 
ence were several emergency strips that 
had been used by Australian civil avi- 
ation before the war. 

Lae, a prewar sea terminal for air serv- 
ice to the Bulolo Valley, had a developed 
harbor and airfield, and was the key to 
successful employment of airfields in the 
valleys. Once it was captured, ships could 
carry supplies to Lae, and roads could 
be pushed up the Markham Valley to 
carry supplies to the airfields. The New 
Guinea Force was ordered to construct 
airfields in the Lae-Markham Valley area 
as specified by General Kenney. They 
were eventually to include facilities for 
two fighter groups, some night fighters, 
two medium and two light bombardment 
groups, one observation squadron, one 
photo-reconnaissance squadron, and four 
transport squadrons, MacArthur wanted 
Madang taken in order to protect the 
Southwest Pacific's left flank during the 
subsequent landings on New Britain. 
Salamaua was not an important objec- 
tive, but MacArthur and Blarney ordered 
the 3d Australian Division with the Mac- 
Kechnie Force attached to press against 
it for purposes of deception. They 
wanted the Japanese to believe that Sala- 
maua and not Lae was the real objective, 
and so to strengthen Salamaua at Lae's 

The commander in chief ordered Ken- 



ney and Carpender to support the New 
Guinea Force with their Allied Air and 
Allied Naval Forces. Allied Land Forces 
could make the necessary troops avail- 

U.S. Army Services of Supply and Line 
of Communications units of Allied Land 
Forces would provide logistical support. 
From thirty to ninety days of various 
classes of supply was to be stocked at in- 
termediate and advance bases. General 
Marshall's U.S. Army Services of Supply 
would be responsible for supply of Amer- 
ican forces in the Huon Peninsula and 
Markham Valley, and would provide all 
items to the Army and Navy. MacArthur 
ordered Marshall's command to aid 
Allied Naval Forces in transporting the 
2d Engineer Special Brigade to the com- 
bat zone, and to prepare to relieve Allied 
Naval Forces of the responsibility for 
transporting supplies to Lae and to 
Woodlark and Kiriwina. 

Some of the plan's outstanding fea- 
tures were the ways it proposed to use 
air power. The impending assault by 
parachute would be the first tactical em- 
ployment of parachute troops as such by 
Allied forces in the Pacific. 2 The com- 
bination of airlifted troops and para- 
chute troops in co-ordination with am- 
phibious assault had also not been used 
hitherto by the Allies in the Pacific. The 
year before, General Whitehead had 
"sold the Aussies on the scheme of an 
airborne show at Nadzab to take Lae out 
from the back," and General MacArthur 
had liked the idea too, but there were 
not enough transport planes in the area 

2 The ist Marine l'arachutc Battalion fought well 
at Guadalcanal and Tulagi in 1942, but it fought 
on foot as an infantry battalion. It made no tactical 

to carry it out at that time. 3 Generals 
MacArthur and Blarney had planned to 
operate overland from the Lakekamu 
River to the Bulolo Valley and thence to 
the Markham Valley in conjunction with 
a parachute assault by one battalion. De- 
lays in building the mountain road from 
the Lakekamu to the Bulolo Valley made 
necessary a decision to land an entire 
parachute regiment at Nadzab, a superb 
airfield site in the Markham Valley where 
a prewar Australian airstrip already 
existed, and to fly an entire division 
from Port Moresby to Nadzab immedi- 
ately afterward. 

The third unusual feature of the Pos- 
tern air operations was made possible by 
General Kenney's enthusiastic willing- 
ness to try any experiment that offered 
a hope of success and by the fact that 
both Allied and Japanese forces were 
concentrated in small enclaves on the 
New Guinea coast, with the highlands 
and hinterland available to whichever 
force could maintain patrols there. On 
General Kenney's recommendation, Mac- 
Arthur ordered the development of two 
grass strips, one in the Watut Valley west 
of Salamaua and the other in the grassy 
plateau south of Madang where the 
Markham and Ramu Rivers rise. These 
strips could serve as staging bases that 
would enable Keaney to send fighters 
from Port Moresby and Dobodura as far 
as the expanding enemy base at Wewak 
or over the western part of New Britain, 
and to give fighter cover to Allied 
bombers in the vicinity of Lae. Thus the 
Allied Air Forces would be using inland 
airfields to support and protect a sea- 
borne invasion of a coastal area. 

3 Kenney, General Kenney Reports, p. 128. 



D Day was set for planning purposes 
as 1 August, but was postponed to 27 
August and finally to 4 September to 
permit the assembly of enough C-47's, 
more training for the 7th Australian Di- 
vision, and the relief of the VII Am- 
phibious Force of its responsibilities for 
Woodlark and Kiriwina. The precise 
date was picked by General Kenney on 
the basis of weather forecasts. He wanted 
fog over western New Britain and Vitiaz 
and Dampier Straits that would keep 
Japanese aircraft away while bright clear 
weather over New Guinea— a fairly com- 
mon condition— permitted the flight to 
and jump into the Markham Valley. The 
fourth of September promised to be such 
a date and was selected.* 

The final tactical plans were prepared 
by New Guinea Force and by the various 
higher headquarters in the Allied Air 
and Allied Naval Forces under the super- 
vision of General MacArthur, General 
Sutherland, MacArthur's chief of staff, 
and such subordinates as General Cham- 
berlin, the G-3 of GHQ. S 

1 In his book Kenney tells how the American and 
Australian weather teams kept altering their fore- 
casts and disagreeing with one another. Finally the 
American team picked 5 September; the Australians 
decided on 3 September. General Kenney "decided 
that neither one of them knew anything about 
■weather, split the difference between the two fore- 
casts, and told General MacArthur we would be 
ready to go on the morning of the 4th for the 
amphibious movement. . . ." General Kenney Re- 
ports, p. 288. 

GHQ supervised the preparation of the plans 
for Operation II more closely than, for example, 
those for Woodlark-Kiriwina. The staff at GHQ 
felt that New Guinea Force and subordinate head- 
quarters were slow in preparing plans, tended to 
prepare plans for initiating operations rather than 
for carrying them through completely, failed to pro- 
vide for co-ordination of forces, and did not 
thoroughly appreciate logistics. See Memo, Cham- 
berlin for CofS GHQ SWPA, 28 Aug 43, no sub, in 
GHQ SWPA G-3 Jnl, 16 Jul 43. 

Final plans, issued in August, called 
for the employment of two veteran Aus- 
tralian divisions, the 7th and the 9th, 
the U.S. 503d Parachute Infantry Regi- 
ment, and elements of the U.S. 2d En- 
gineer Special Brigade, as well as the 
American and Australian troops already 
pressing against Salamaua in their de- 
ception maneuver. The 9th Australian 
Division was to be carried by the VII 
Amphibious Force, with elements of the 
2d Engineer Special Brigade, attached, 
from Milne Bay to beaches far enough 
east of Lae to be beyond range of enemy 

Early plans had called for the 2d En- 
gineer Special Brigade to carry the 9th 
Australian Division to Lae and support 
it thereafter. But closer study showed 
that an engineer special brigade could 
carry and support but one brigade in re- 
duced strength— about 3,000 men, or not 
nearly enough to attack Lae. Therefore 
the VII Amphibious Force was ordered 
to carry the 9th Division, and the 2d En- 
gineer Special Brigade was attached to 
Barbey's command for the initial phases. 
Two brigade groups, totaling 7,800 men, 
plus elements of the amphibian en- 
gineers, were to land starting at 0630, 4 
September. |; That evening 2,400 more 
Australians would land, and on the night 
of 5-6 September the VII Amphibious 
Force, having retired to Buna after un- 
loading on 4 September, was to bring in 
the 3,800 men of the reserve brigade 
group. The time for H Hour, 0630, was 
selected because it came thirty minutes 
past sunrise, by which time the light 
would be suitable for the preliminary 
naval bombardment. 

" A brigade group was similar in strength and 
composition to a U.S. regimental combat team. 



Admiral Carpender organized his Al- 
lied Naval Forces into almost the same 
task forces that he had set up for Wood- 
lark and Kiriwina and assigned them 
similar missions. Admiral Barbey organ- 
ized his VII Amphibious Force into a 
transport group of 2 destroyers, 4 APD's, 
13 LST's, 20 LCI's, 14 LCT's and 1 AP; 
a cover group of 4 destroyers; an escort 
group of 2 destroyers; an APC group of 
13 APC's, g LST's, and 2 subchasers; and 
a service group of 1 tender, 3 LST's, 10 
subchasers, 5 minesweepers, 1 oiler, and 
1 tug. The attached engineer special bri- 
gade elements possessed 10 LCM's and 40 

Allied Air Forces' plans for support of 
the invasion called for General White- 
head to provide close support to ground 
troops, to provide escort and cover for 
the amphibious movements, to establish 
an air blockade over Huon Peninsula, 
to specify to General Blarney the air fa- 
cilities to be constructed in the target 
areas, and to prepare to move forward 
to the new bases. 

But again there was an argument over 
the method by which the air forces would 
cover the VII Amphibious Force. Ad- 
mirals Carpender and Barbey had no air- 
craft carriers and thus were completely 
dependent upon the Allied Air Forces 
for air support. They pointed out that 
the amphibious movement to Lae would 
involve over forty ships, 7,800 soldiers 
and 3,260 sailors. This represented all 
suitable vessels available, with none re- 
tained in reserve. Losses to Japanese air 
attacks would seriously jeopardize the 
success of future operations, and there- 
fore they argued that only a fighter um- 
brella providing continuous cover for 
the VII Amphibious Force would be 

adequate. The airmen, who were plan- 
ning to use over three hundred planes 
in the Markham Valley parachute jump, 
were willing to provide air cover for 
Barbey's ships over Lae itself on D Day, 
but argued that the movement of the 
convoys would be amply protected by 
maintaining fighter squadrons on ground 
alert at Dobodura and the staging air- 
field in the Watut Valley. The argument, 
a heated one, went up the chain of com- 
mand to General MacArthur himself, 
and was finally settled by Kenney's agree- 
ment to use a total of thirty-two planes 
to give as much cover as possible over 
the VII Amphibious Force during day- 
light and to maintain fighter squadrons 
on ground alert. 

There remained the problem of fighter 
control. One fighter control unit was 
stationed at Dobodura, and another at 
the staging field in the Watut Valley, but 
radar coverage over the area was far from 
complete. Japanese aircraft from Wewak 
or Madang could fly south of the moun- 
tains to Lae, or from New Britain across 
Dampier and Vitiaz Straits, and radar 
would not pick them up until they were 
almost over Lae. And as Brig. Gen. 
Charles A. Willoughby, MacArthur's 
G-2, pointed out, Allied experience at 
New Georgia showed that the Japanese 
air reaction might be violent. An Aus- 
tralian airman suggested that the dif- 
ficulty be alleviated by posting a radar- 
equipped destroyer between Lae and 
Finschhafen. This was accepted, and the 
U.S. destroyer Reid, which was part of 
Barbey's antisubmarine screen, was se- 
lected as picket with orders to steam in 
Vitiaz Strait some forty-five miles south- 
east of Finschhafen. 

Markham Valley plans called for the 



503d Parachute Infantry Regiment, fly- 
ing from Port Moresby in C-47's, to 
jump onto Nadzab airstrip on the north 
bank, of the Markham River on 5 Sep- 
tember, the day after the amphibious as- 
sault. Nadzab was not believed to be 
occupied by the Japanese, but this seizure 
would block the valley and prevent the 
enemy's sending troops overland from 
Wewak. Once captured, Nadzab airstrip 
was to be quickly readied for airplanes 
by the 503d and by a force of Australian 
engineers and pioneers. The Australians 
were to paddle in boats from the staging 
airfield in the Watut Valley down the 
Watut River to Nadzab— a distance of 
about thirty-two air miles, but actually 
twice that far for anything but crows and 
airplanes. Then one brigade of the 7th 
Australian Division, plus engineers and 
antiaircraft units, having been flown to 
the Watut Valley previously, would fly 
in. The next brigade would come in di- 
rectly by air from Port Moresby. Once 
adequate strength had been assembled, 
the 7th Australian Division would march 
eastward down the Markham River 
against Lae, and at the same time the 9th 
Australian Division would drive west- 
ward from the landing beaches. 

Seizure of Nadzab would have a three- 
fold effect: it would provide Allied forces 
with one more air base with which to 
increase their control over the Huon 
Peninsula, the straits, and western New 
Britain; it would provide a base for the 
7th Division's eastward march against 
Lae; and an Allied force at Nadzab 
could forestall any attempt by the Japa- 
nese to reinforce Lae from Wewak by 
marching through the Ramu and Mark- 
ham Valleys. 

The Enemy 

Japanese strategic intentions were not 
changed by the invasion of the Trobri- 
ands or of Nassau Bay. In August 1943 
Generals Imamura and Adachi were still 
resolved to hold Lae and Salamaua as 
parts of the outer defenses of Wewak 
and Madang, and were still planning to 
move into Bena Bena south of the Ramu 
Valley. 7 There were about ten thousand 
men in the Lae-Salamaua area, with 
somewhat more than half of these de- 
fending Salamaua. Many of the ten thou- 
sand, reported the Japanese after the 
war, were sick. Some estimates run as 
high as 50 percent. At Lae, General 
Shoge, temporarily detached from his 
post as infantry group commander of the 
41st Division, led a force consisting of a 
naval guard unit, elements of the 21st, 
i02d, and 115th Infantry Regiments, 
and artillerymen and engineers. In ad- 
dition to defending Lae, Shoge was re- 
sponsible for patrolling up the Markham 
River and for protecting the southern 
approaches to Finschhafen on the east 
coast of the Huon Peninsula. 

In the months following the Bismarck 
Sea disaster the supply systems for Lae 
and Salamaua had almost broken down. 
The Allied aerial blockade of the Huon 
Peninsula prevented the use of large 
ships to carry supplies forward to Lae. 
Until June, six submarines helped carry 
supplies, but then the number was cut 
to three and the bulk of supplies had to 
be carried on barges. Supply of the ten 

' This subsection is based upon 8th Area Army 
Operations, Japanese Monogr No. 110 (OCMH), 
pp. 23-34, 36-85; 18th Army Operations, II, Japa- 
nese Monogr No. 43 (OCMH), 27-54; USSBS, Allied 

Campaign Against Rabaul, p. 84. 



thousand men for the five months pre- 
ceding September would have required 
150 bargeloads per month, while 200 
more barges were needed for transport 
of reinforcements and ammunition. But 
there were far too few barges. Only 40, 
for example, were making the run to 
Lae from the staging base at Tuluvu on 
the north shore of Cape Gloucester. The 
sea and the tides in Dampier Strait dam- 
aged many, and several fell victim to 
Allied aircraft and to nocturnal PT's 
which, like their sister boats in the 
Solomons, prowled the barge lanes. 

Imperial General Headquarters, mean- 
while, had paid heed to Imamura's re- 
quest for more planes. On 27 July Im- 
perial Headquarters ordered the 4th Air 
Army, commanded by Lt. Gen. Kumai- 
chi Teramoto, from the Netherlands 
Indies to the Southeast Area. Teramoto's 
army would include the jth Air Division, 
the 14th Air Brigade, some miscellane- 
ous squadrons, and the 6th A ir Division, 
which was already based at Wewak. 

The 4th Air Army headquarters ar- 
rived at Rabaul on 6 August, whereupon 
Imamura ordered Teramoto and his 
planes to proceed to Wewak with the 
mission of escorting convoys, destroying 
Allied planes and ships, and co-operat- 
ing with the 18 th Army. The move was 
made at once; the Allies were well aware 
that the Japanese were building up 
strength on the four Wewak airfields. 

Allied Air and Naval Preparations 
Increases in Air Strength 

The increases in Allied strength that 
had been promised to the Southwest 
Pacific Area at the Pacific Military Con- 
ference in March had been coming 

through practically on schedule. 8 P-47's 
of the 348th Fighter Group began arriv- 
ing in Australia in June, and before the 
end of July the whole group had been 
deployed to New Guinea. The 475th 
Fighter Group, flying P-38's, was ready 
for combat by the middle of the next 

Bomber strength, too, was increasing. 
Newly arrived B-24's of the 380th Heavy 
Bombardment Group went into action 
from Darwin, Australia, in mid-July. 
One of the 38oth's first large-scale opera- 
tions was a spectacular raid on the oil 
center at Balikpapan, Borneo, on 13 
August, a feat that required a 1,200-mile 
round trip. Port Moresby saw the arrival 
of new B-25's of the 345th Medium 
Bombardment Group in July. And the 
C-47's were also increasing in number. 
By September the 54th Troop Carrier 
Wing could boast fourteen full squad- 
rons of transport planes. 

By the end of August the Southwest 
Pacific Area had on hand nearly all its 
authorized plane strength— 197 heavy 
bombers and 598 fighters. Keeping this 
number in flying condition, however, 
was next to impossible. Many of the 
planes were old, and with the air forces 
constantly in action there were always 
battle casualties. Kenney was always 
short of manpower; he could never ob- 
tain enough replacement pilots to keep 
all his new and veteran squadrons up 
to strength, a condition that was prob- 
ably duplicated in every active theater. 


The first important action of Kenney's 

* This section is based on Craven and Cate, The 
Pacific: Guadalcanal to Saipan, pp. 168-86; Mori- 
son, Breaking the Bismarcks Barrier, pp. 257-61; 
Kenney, General Kenney Reports, pp. 251—79. 



Allied Air Forces in preparation for the 
Markham Valley-Huon Peninsula opera- 
tion was the development of the staging 
fields in the Watut Valley and in the 
Ramu-Markham trough. Ever since the 
Buna campaign Kenney had been anx- 
ious for a good fighter field near Lae to 
use in covering the invasion. He hoped 
to fly troops into an existing emergency 
strip and seize it, as he had done during 
the Buna campaign. Kokoda and Wau 
had been surveyed but found unsuitable. 
Then in May an aviation engineer of- 
ficer, with orders to find a field farther 
forward than Wau, trekked from the 
Bulolo Valey almost to Salamaua, found 
nothing suitable, and thereupon back- 
tracked and went down the Watut River 
where he found and recommended an 
emergency landing strip at Marilinan. 
But Marilinan was not perfect; it was 
feared the September rains would render 
its clay too muddy to be usable. At this 
point General Wurtsmith of the V 
Fighter Command took a hand. Look- 
ing over the ground himself, he picked 
a site at Tsili Tsili four miles down the 
Watut River from Marilinan. Kenney 
and Whitehead agreed with his choice. 

Meanwhile Kenney and Herring ar- 
ranged to build the second staging field, 
using a few Australian troops and native 
labor, at Bena Bena south of the Ramu 
Valley. This emergency strip had long 
served as a New Guinea Force patrol 
base, and the Japanese at this time were 
hoping to capture it eventually. The 
Allies decided to build a grass strip suit- 
able for fighters at Bena Bena (C-47's 
carrying supplies to the Australian pa- 
trols had been using Bena Bena for 
some time), and to burn off the grass in 
fashion so obvious as to distract the 

enemy's attention from Tsili Tsili. 

In June and July, C-47's flew Aus- 
tralian troops and the U.S. 871st Air- 
borne Engineer Battalion to Marilinan. 
The troops moved down the river to 
Tsili Tsili, cleared the strips, and C-47's 
flew in specially designed bulldozers and 
other earth-moving equipment. Some 
gear, including trucks sawed in half so 
they could be loaded into C-47's, was 
also flown to Tsili Tsili, where the trucks 
were welded together. Two strips at 
Tsili Tsili were soon ready, and by mid- 
August three thousand troops, including 
a fighter squadron, were based there. 
Japanese aircraft failed to molest the 
Allies until the fields were all built; they 
raided Tsili Tsili on 15 and 16 August 
without doing much damage and there- 
after left it alone. 

While General Kenney had liked the 
prospects of Tsili Tsili from a technical 
point of view, he had felt that Tsili 
Tsili had an unfortunate sound. He 
therefore officially directed that the base 
be given the more attractive name of 
Marilinan. 9 

During this period the Fifth Air Force 
had been supporting the Allied diver- 
sionary attacks against Salamaua. Nearly 
every day of July saw some form of air 
attack against the Lae-Salamaua area. 
Sorties during the month totaled 400 by 
B-25's, 100 by B-24's, 45 by RAAF Bos- 
tons, 35 by A-20's, 30 by B-i7's, and 7 
by B-26's. The Japanese supply point at 
Madang was also raided during the 
period 20-23 July by B-25's and heavy 

But these raids were secondary to 
Kenney's main air effort, which was di- 

9 Craven and Cate in The Pacific: Guadalcanal to 
Saipan, however, use the name Tsili Tsili. 

B-24 Over Salamaua, 13 August 1943. Note smoke from bomb bursts. 



Enemy Aircraft Destroyed on the Ground by Allied planes near Lae. 

rected against Wewak. Aware of the in- 
crease in Japanese air strength at Wewak, 
and lacking enough strength to hit both 
Wewak and Rabaul, Kenney had decided 
to concentrate against Wewak rather 
than Rabaul up to the day of the land- 
ing at Lae, and to rely in part on the 
weather for protection against Rabaul- 
based planes. There were too many Japa- 
nese fighter planes at Wewak, however, 
for Kenney to risk sending unescorted 
bombers there. Raids against Wewak had 
to await completion of the Marilinan 
staging field, which would extend the 
range of Allied fighters as far as Wewak. 
Meanwhile Kenney ordered his bombers 
not to go as far as Wewak, thus leading 

the Japanese to believe that Wewak lay 
beyond bomber range and to send planes 
there with a false sense of security. 

On 13 August photographs taken by 
Allied reconnaissance planes showed a 
total of 199 Japanese airplanes on the 
four fields at Wewak. The 4th Air Army 
was now due for a surprise. Marilinan 
was ready by midmonth and so was the 
Fifth Air Force. General Whitehead had 
four bombardment groups with enough 
range to hit Wewak from Port Moresby 
—two heavy groups with 64 planes in 
commission and two medium groups 
totaling 58 B-25's. With Marilinan in 
commission the bombers would have 
fighter protection all the way. 



B-25 Medium Bombers leaving installations aflame in the Wewak area. 

Heavy and medium bombers and 
fighters struck the four Wewak fields on 
17 August and achieved excellent results. 
Taking the Japanese by surprise, they 
caught most of the enemy planes on the 
ground. Next day they were back in 
strength, and the "Wewak offensive con- 
tinued throughout the rest of August. 
The planes struck at Hansa Bay and 
Alexishafen during the same period. 

Damage inflicted by these raids was 
heavy, though less than estimated at the 
time. Kenney's headquarters claimed 
over 200 Japanese aircraft destroyed on 
the ground, a claim that Army Air 
Forces headquarters scaled down to 175. 
Postwar Japanese reports, however, give 

losses as about half what the Allies 
initially claimed. 10 But despite the efforts 
of Imamura and Teramoto, strength of 
the 4th Air Army thereafter averaged 
but 100 planes, and "the prospect of the 
New Guinea operation [was] much 
gloomier." 11 

The Allied Naval Forces, which had 
not played a decisive part in the Buna 
campaign because it lacked enough ships 

1(> 8th Area Array Operations, Japanese Monogr 
No. 110 (OCMH), p. 83, states that one hundred 
planes were lost; 18th Army Operations, II, Japa- 
nese Monogr No. 4a (OCMH), zg, asserts that sixty 
to seventy were destroyed. 

" 8th Area Army Operations, Japanese Monogr 
No. 110 (OCMH), p. 84. 



and because hydrographic information 
on the waters of New Guinea's north 
coast was almost nonexistent, was also 
taking a hand. PT boats based at Morobe 
were stalking the enemy barge routes at 
night and making the transport of men 
and munitions to Lae increasingly dif- 
ficult. The Fifth Air Force's successful 
strike against Wewak encouraged Ad- 
miral Carpender to send warships as far 
up the coast as Finschhafen. Thus on 22 
August four destroyers under Capt. Jesse 
H. Carter left Milne Bay, stopped at 
Buna to discuss air cover and obtain 
target information, and sailed for Finsch- 
hafen. Starting at 0121, 23 August, 
Carter's ships bombarded Finschhafen 
with 540 rounds of 5-inch shells and re- 
turned safely to Milne Bay. This opera- 
tion was small in itself, but it was sig- 
nificant because this was the first time 
Allied warships had ventured so far up 
the New Guinea coast. 

During the first three days of Septem- 
ber Allied planes executed preparatory 
bombardments in support of the Lae in- 
vasion. They launched heavy attacks 
against airfields, supply points, and ship- 
ping lanes on 1 September, the same day 
on which medium and heavy bombers 
raided Alexishafen and Madang. Next 
day B-25's and P-38's delivered a low- 
level attack against Wewak. Gasmata and 
Borgen Bay on New Britain, and Lae it- 
self, were struck on 3 September, and 
eleven nocturnal RAAF Catalinas raided 

The Salamaua Attack, 
1 July-12 September 1943 

During July and August, while the 
various headquarters of the Southwest 

Pacific Area were preparing plans and 
assembling troops and supplies for the 
Lae-Markham Valley invasions, and 
while the air and naval forces were at- 
tacking Japanese aircraft, bases, and lines 
of communication, the troops in front of 
Salamaua were carrying out their part 
of the plan by the diversionary attack 
against that port. 12 Starting from the 
arc-shaped positions they held in early 
July, the 3d Australian Division and the 
MacKechnie Force, soon joined by the 
remainder of the i62d Infantry, fought 
their way forward until by the end of 
August they were closing i n on the town 

and airfield of Salamaua. | (See Map 6.) 

At first the reinforced 1st Battalion, 
i62d Infantry, fighting on the right of 
the 3d Australian Division, was the only 
American unit present, but this force 
was enlarged in July, after the capture of 
Bitoi Ridge, when other elements of the 
41st Division were attached to the 3d 
Australian Division. This attachment 
came about because more U.S. infantry- 
men and artillerymen were needed to 
secure the Tambu Bay-Dot Inlet area 
northwest of Nassau Bay, and because a 
supply base for Australians and Ameri- 
cans in the combat area was required. 
Consequently the Coane Force, com- 
manded by Brig. Gen. Ralph W. Coane 
who was also 41st Division artillery com- 
mander, was organized during the second 

11 Unless otherwise indicated this section is based 
on McCartney, The Jungleers, Ch. VII; ALF, Rpt 
on New Guinea Opns: 1 Mar— 13 Sep 43; i62d Inf 
Rpt of Opns, and Jnl; 41st Div Arty, Hist of 
Salamaua Campaign, 23 Apr-4 Oct 43; Combined 
Operational Int Center GHQ SWPA, Resume of 
Allied Mil Opns and Int Leading to the Capture 
of Lae and Salamaua From the Enemy: Jun-Sep 43, 
20 Sep 43, in GHQ SWPA Jnl, so Sep 43. 



week in July. 13 The MacKechnie Force, 
then fighting forward from Bitoi Ridge, 
was not a part of the Coane Force. Some 
units assigned to the Coane Force were 
already in the Nassau Bay area; others 
soon came up from Morobe. 14 

Both Coane and MacKechnie Forces 
fought under command of General 
Savige, commanding the 3d Australian 
Division, and after Savige's headquarters 
was relieved on 24 August by Head- 
quarters, 5th Australian Division, under 
command of Maj. Gen. E. J. Milford, the 
Americans served under Milford. At the 
same time Col. William D. Jackson, 41st 
Division artillery executive officer, was 
appointed as Commander, Royal Artil- 
lery, of the 3d and then the 5th Aus- 
tralian Divisions. Jackson, using a com- 

13 It consisted at first of the ad and 3d Battalions, 
t6ad Infantry; the i6ad Infantry Cannon Company; 
3d Platoon, Antitank Company, i6ad Infantry; C 
Battery, 209th Coast Artillery Battalion; A Battery, 
218th Field Artillery Battalion; A Company, 116th 
Medical Battalion; A and D Companies, 532(1 En- 
gineer Boat and Shore Regiment, 2d Engineer 
Special Brigade; A Company, Papuan Infantry Bat- 
talion; a Combined Operational Service Command 
detachment; Troop D, 2/6th Royal Australian 
Artillery Regiment; and signal and quartermaster 

" Problems involving command over the mixed 
Australian-American units appear to have been ad- 
ditional factors in the decision to create the Coane 
Force rather than to turn all American troops over 
to Colonel MacKechnie. The MacKechnie Force was 
attached to the 3d Australian Division, but Maj. 
Gen. Horace H. Fuller, commanding the 41st U.S. 
Division, retained control over the American troops 
at the actual beachhead. Thus, as he put it later, 
Colonel MacKechnie was "placed in the unenviable 
position of trying to obey two masters" who kept 
giving him conflicting orders. The impossibility of 
obeying them both finally led MacKechnie to re- 
quest relief as commanding officer of the lfiad. He 
was reassigned as Coane Force S— g, and later as 
liaison officer with the 3d Australian Division, but 
returned to command the i62d on the dissolution 
of the Coane Force. See Ltr, Col MacKechnie to 
Gen Smith, Chief of Mil Hist, 2 Nov 53, no sub, 

posite Australian and American staff, 
served as artillery commander until the 
end of hostilities in that area. 

On 17 July the Coane Force moved 
forward from Nassau Bay, and by the 
end of the next day had secured the 
southern headland of Tambu Bay, where 
a supply base was set up. Starting on 20 
July, the Americans launched a series of 
attacks with strong artillery support 
which resulted on 13 August in the cap- 
ture of the high ground— Roosevelt 
Ridge, Scout Ridge, and Mount Tambu 
—overlooking Tambu Bay and Dot Inlet. 
On 12 August the Coane Force was dis- 
solved and the entire i62d Infantry re- 
verted to Colonel MacKechnie's control. 

At the same time the Australians 
pressed forward so that by the first week 
in September they had reached the Fran- 
cisco River, which flows in an west-east 
direction just south of the Salamaua air- 
field. All advances were made up and 
down precipitous ridges varying from 
eight hundred to three thousand feet in 
height. With characteristic skill the Jap- 
anese had established strong defensive 
positions on the ridges; there were many 
automatic weapons emplacements, with 
earth-and-log pillboxes predominating, 
that gave each other mutual support with 
interlocking bands of fire. Trenches and 
tunnels connected the emplacements. 

Early September saw Japanese resist- 
ance slackening. On 1 1 September the 
Australians and the i62d Infantry Re- 
connaissance Platoon crossed the rain- 
swollen Francisco River and by the end 
of 12 September the airfield, the town, 
and the entire isthmus, w r hich had been 
held by the Japanese for eighteen 
months, was back in Allied hands. 

The cost was not cheap. On 29 June 

Salamaua, objective of the attack. 

there were 2,554 men in the 162c! In- 
fantry. By 12 September battle casualties 
and disease had reduced the regiment to 
1,763 men. One hundred and two had 
been killed, 447 wounded. The 163d 
estimated it had killed 1,272 Japanese 
and reported the capture of 6 prisoners. 

The Japanese had lost Salamaua after 
a stiff fight and the very strength of their 
defense had played into Allied hands, 
for of the ten thousand enemy soldiers 

in the Lae-Salamaua area, the majority 
had been moved to Salamaua. The Allied 
ruse had succeeded. 

Lae: The Seaborne Invasion 

The Landing 

The unit slated to invade Lae, Maj. 
Gen. G. F. Wootten's 9th Australian Di- 
vision, embarked on the ships of Ad- 
miral Barbey's Task Force 76 at Milne 



Crossing Rain-Swollen Francisco River 

Bay on 1 September. 15 Next day Barbey's 
ships sailed to Buna and to Morobe, 
where they were joined by fifty-seven 
landing craft of the 2d Engineer Special 

10 This section is based on Morison, Breaking the 
Bismarcks Barrier, Ch. XIV; Off of Chief Engr, 
GHQ AFPAC, Engineers in Theater Operations, p, 
106, and Critique, p. 97; ALF, Rpt on New Guinea 
Opns: 4 Sep 43-26 Apr 44; Combined Operational 
Int Center GHQ SWPA, Resume . . . Lae and 
Salamaua, in GHQ SWPA G-g Jnl, 30 Sep 43; 2d 
ESB Rpt. 

Brigade that had assembled there in the 
latter part of August. On the night of 
3-4 September the armada set out for 
Lae, eighty miles distant; it arrived at 
the landing beaches east of L ae at sun- 

rise of 4 September. (Map 13) 

At 0618, eighteen minutes after the 
sun rose, five destroyers fired a ten- 
minute bombardment on the beaches. 
Then sixteen landing craft from the 
APD's started for the beaches carrying 



503 t+J 


; 4-16 September 1943 




^ 1 1 I I 

MAP 13 

the assault waves. At 0631 the 20th Aus- 
tralian Infantry Brigade began going 
ashore at Red Beach, near Bulu Planta- 
tion and some eighteen miles east of Lae. 
This landing was unopposed. Two 
minutes later troops of the 26th Aus- 
tralian Infantry Brigade landed at 
Yellow Beach, eighteen miles east of 
Lae, east of the Bulu River. A small 
group of Japanese on Yellow Beach ran 
away at the approach of the Australians. 
Scouts of the 2d Engineer Special Bri- 
gade landed with the Australian in- 

Fifteen minutes after the assault waves 
beached, LCI's pushed their bows onto 
the beaches and put more riflemen 
ashore. They were followed by LCT's 
and LST's. All assault troops had landed 
by 0830, and by 1030 fifteen hundred 
tons of supplies had been landed. By 
the end of the day the beachheads were 
secure, 2,400 more Australians had 
landed, and the 26th Brigade and the 
2/ 17th Australian Infantry Battalion had 
crossed the Buso and begun the advance 
westward against Lae. 

There was no resistance on the 


V. Holmes ,Jr. 

ground, but Japanese aircraft attempted 
to break up the invasion. About 0700, 
before fighter cover had arrived, a few 
two-engine bombers with fighter escort 
attacked Task Force 76 and damaged two 
LCI's. Imamura dispatched eighty planes 
from Rabaul to attack Barbey but these 
were delayed by the fog over New 
Britain that Kenney's weathermen had 
predicted. The picket destroyer Reid's 
radar located them over Gasmata in the 
afternoon just as Task Force 76 was 
making ready to sail for Milne Bay. The 
Reid vectored out forty P-38's and 

twenty P~47's which intercepted the 
flight and broke it up. Some planes got 
through, however, and attacked a group 
of six LST's off Cape Ward Hunt. They 
damaged two and killed over a hundred 
Australian soldiers and American sailors. 
The Japanese did not attack the jammed 
landing beaches at this time, but re- 
turned in the evening to blow up an am- 
munition dump, damage two beached 
LCI's, and kill two men." 

"They claimed to have sunk 14 transports, a 
barges, 1 PT boat, 3 destroyers, and to have shot 
down 38 planes. 

The Advance Westward 

Once the assault troops had landed 
control of the 2d Engineer Special Bri- 
gade elements— thirteen hundred men of 
a reinforced boat company, a boat con- 
trol section, a shore battalion, a medical 
detachment, scouts, and a headquarters 
detachment— passed from Admiral Bar- 
bey to General Wootten. A salvage boat, 
ten LCVP's, and two additional LCVP's 
mounting machine guns for support of 
landings remained at Red Beach. Even- 
tually, twenty-one LCM's and twenty- 
one LCVP's were sent to Red Beach. 
Because of breakdowns, these replace- 
ments were necessary if ten craft of each 
type were to be kept in operation. All 
these craft were used to support the gth 
Division's march against Lae. 

The 2/ 13th Australian Infantry Bat- 

talion, once landed, pushed east from 
Bulu Plantation and secured the east 
flank by seizing Hopoi. The reserve 24th 
Infantry Brigade landed on schedule on 
the night of 5-6 September, and at day- 
light started west behind the 26th Bri- 
gade. On 6 September, after a ten-mile 
march, the 26th Brigade met its first op- 
position at the Bunga River. 

The 24th Brigade advanced along the 
coast while the 26th Brigade moved some 
distance inland in an effort to get be- 
hind Lae and cut off the enemy garrison. 
The 24th's advance was rendered dif- 
ficult, not so much by the enemy as by 
the terrain. The heavy September rains 
flooded the creeks and turned the trails 
into deep mud that was virtually im- 
passable for vehicles. Fortunately the 
boats of the 2d Engineer Special Brigade 



were available to ferry supplies by water 
to coastal dumps and enable the advance 
to continue. The leading Australian bat- 
talions reached the Busu (not to be con- 
fused with the Buso farther east) on the 
morning of 8 September. This swollen 
river, five feet deep and sixty feet wide 
at the mouth, and flowing at twelve 
knots, was a severe obstacle in itself, and 
the west bank was held by the Japanese. 

Patrols attempted to force a crossing 
on the morning of 9 September but the 
combination of Japanese bullets and the 
swift current forced them back. In the 
late afternoon elements of four rifle com- 
panies got across in rubber boats and by 
wading and swimming. Several men 
were drowned and many weapons lost in 
this act of gallantry, but the four com- 
panies seized a bridgehead on the west 
bank and held it against enemy counter- 

Meanwhile the troops on the east 
bank loaded men, weapons, and am- 
munition onto the 2d Engineer Special 
Brigade's landing craft and sent them to 
the west bank. For the next sixty hours, 
the landing craft plied back and forth 
until the entire 24th Brigade had been 
transferred to the west bank. Rain, mist, 
and darkness helped hide the boats from 
the Japanese, who tried to hit them with 
artillery, machine guns, and rifles. Dur- 
ing the same period a box girder bridge 
was moved in pieces by landing craft 
from Bulu Plantation to the mouth of 
the Burep River, then laboriously hauled 
inland to the 26th Brigade's zone over a 
jeep track built by the 2d Engineer Spe- 
cial Brigade. The bridge was installed 
over the Busu under enemy fire on the 
morning of the 14th. The 26th Brigade 
crossed over that night. Both brigades 

were then on the west bank of the Busu 
and were ready to resume the advance 
against Lae and effect a junction with 
the troops of the 7th Australian Division 
that were advancing east out of Nadzab. 

Nadzab: The Airborne Invasion 

The Jump 

Capture of Nadzab had been spectac- 
ularly effected on 5 September. This mis- 
sion, assigned to Col. Kenneth H. Kins- 
ler's 503d Parachute Infantry Regiment, 
was coupled with the additional mission 
of preparing the airstrip for C-47's carry- 
ing Maj. Gen. George A. Vasey's 7th 
Australian Division from Marilinan and 
Port Moresby. 17 

Reveille for the men of the 503d 
sounded early at Port Moresby on the 
morning of 5 September. The weather 
promised to be fair, although bad flying 
weather over the Owen Stanleys delayed 
take off until 0825. New Guinea Force 
had prepared its plans flexibly so that 
the seaborne invasion on 4 September 
would not be slowed or altered if any 
threat of bad weather on 5 September de- 
layed the parachute jump, but Kenney's 
weathermen had forecast accurately. 

The paratroopers and a detachment of 

" Unless otherwise indicated this section is based 
on Craven and Cate, The Pacific: Guadalcanal to 
Saipan; pp. 184-86; Morison, Breaking the Bis- 
marchs Barrier, pp, 266-69; Kenney, General Ken- 
ney Reports, pp. 2g2-g6; ALF, Rpt on New Guinea 
Opns: 4 Sep 43—26 Apr 44; Combined Operational 
Int Center GHQ SWPA, Resume . . . Lae and 
Salamaua, in GHQ SWPA G-3 Jnl, 20 Sep 43; 503d 
Parachute Inf Rpt of Opns, Markham Valley, 5—19 
Sep 43; Japanese Operations in the Southwest 
Pacific Area, MS (Vol. II of the MacArthur hist), 
Ch. VII, OCMH; 8th Area Army Operations, Japa- 
nese Monogr No. 110 (OCMH), pp. 81—85; 18th 
Army Operations, II, Japanese Monogr No. 42 
(OCMH), 18-80; Interrogation of Adachi et al., by 
Mil Hist Sec, Australian Army Hq, OCMH. 



C-47 Transport Planes Loaded With Parachute Troops for the drop at 
Nadzab. Two men at left are General Kenney and General MacArthur. 

2/4-th Australian Field Regiment which 
was to jump with its 25-pounder guns 
reached the airfield two hours before 
take off. 18 There they put on parachutes 
and equipment. The 54th Troop Carrier 
Wing had ninety-six C-47's ready, and 
the troops boarded these fifteen minutes 
before take-off time. 

The first C-47 roared down the run- 
way at 0825; by 0840 all transports were 
aloft. They crossed the Owen Stanleys, 
then organized into three battalion 
flights abreast, with each flight in six- 
plane elements in step-up right echelon. 

An hour later bombers, fighters, and 
weather planes joined the formation over 
Marilinan, on time to the minute. All 
together 302 aircraft from eight different 
fields were involved. The air armada 

"The 503d had trained this detachment, 

then flew down the Watut Valley, swung 
to the right over the Markham River, 
and headed for Nadzab, The C-47's 
dropped from 3,000 feet to 400-500 feet. 
The parachutists had stood in their 
planes and checked their equipment over 
Marilinan, and twelve minutes later they 
formed by the plane doors ready to 

In the lead six squadrons of B-25 
strafers with eight .50-caliber machine 
guns in their noses and six parachute 
fragmentation bombs in their bays 
worked over the Nadzab field. Six A-20's 
laid smoke after the last bomb had ex- 
ploded. Then came the C-47's, closely 
covered by fighters. 

The paratroopers began jumping from 
the three columns of C-47's onto separate 
jump areas about 1020. Eighty-one 
C-47's carrying the 503d were emptied 



Airdrop at Nadzab, Morning of 5 September 1943. The paratroopers began 
jumping from C—^j's onto separate jump areas about 1020. 

in four and one-half minutes. All men 
of the 503d but one, who fainted while 
getting ready, left the planes. Two men 
were killed instantly when their chutes 
failed to open, and a third landed in a 
tree, fell sixty feet to the ground, and 
died. Thirty-three men were injured. 
There was no opposition from the 
enemy, either on the ground or in the 
air. Once they reached the ground, the 
503d battalions laboriously moved 
through high kunai grass from landing 
grounds to assembly areas. 

Five B-17's carrying supply parachutes 
stayed over Nadzab all day. They 
dropped a total of fifteen tons of sup- 
plies on ground panel signals laid by the 
503d. The Australian artillerymen and 
their guns parachuted down in the after- 
noon. The whole splendid sight was wit- 
nessed by Generals MacArthur and Ken- 

ney from what Kenney called a "brass- 
hat" flight of three B-17's high above. 
MacArthur was in one, Kenney in an- 
other, and the third B-17 was there to 
provide added fire power in case the 
Japanese turned up. 

The sogd's 1st Battalion seized the 
Nadzab airstrip and began to prepare it 
to receive C-47's, The 2d and 3d Bat- 
talions blocked the approaches from the 
north and east. As soon as the para- 
chutists had begun landing, the Aus- 
tralian units that had come down the 
Watut River— the 2/2d Pioneer Bat- 
talion, the 2/6th Field Company, and 
one company of the Papuan Infantry 
Battalion— began landing on the north 
bank of the Markham. They made con- 
tact with the 503d in late afternoon and 
worked through the night in preparing 
the airstrip. 



The next morning the first C-47 ar- 
rived. It brought in advance elements of 
the U.S. 871st Airborne Engineer Bat- 

Twenty-four hours later C-47's 
brought in General Vasey's 7th Division 
headquarters and part of the 25 th Aus- 
tralian Infantry Brigade Group from 
Marilinan, where they had staged from 
Port Moresby. Thereafter the transports 
flew the Australian infantry and the 
American engineers directly from Port 
Moresby. By 10 September the well- 
timed, smoothly run operation had pro- 
ceeded fast enough that 7th Division 
troops at Nadzab were able to relieve 
the 503d of its defensive missions. 
Enough American engineers had arrived 
to take over construction of new air- 

The sogd's only contact with the 
enemy came in mid-September when 
the 3d Battalion ran into a Japanese 
column at Yalu, east of Nadzab. The 
parachute regiment was withdrawn on 
17 September. It had lost 3 men killed 
jumping, 8 men killed by enemy action, 
33 injured jumping, 12 wounded by the 
enemy, and 26 sick. 19 

This was, comparatively, small cost 
for the seizure of a major airbase with a 
parachute jump. Nadzab paid rich divi- 
dends. Within two weeks the engineers 
had completed two parallel airstrips six 
thousand feet long and had started six 

The Advance Against Lae 
The 25th Australian Infantry Brigade 

" These figures are taken from a table of casualties 
attached to the 503d's report and differ slightly 
from casualty figures in the body of the report. 

Group moved eastward out of Nadzab 
toward Lae on 10 September while Gen- 
eral Wootten's gth Division troops were 
forcing a crossing over the Busu River 
east of Lae. The Markham Valley nar- 
rows near Lae, with the Atzera Range 
on the northeast and the wide river on 
the southwest. A prewar road in the 
Atzera foothills connected Nadzab with 
Lae, and a rough trail on the other side 
of the Atzeras paralleled this road from 
Lae to Yalu, where it intersected the 
road. Thus while some troops blocked 
the trail at Yalu, and the 2/33d Aus- 
tralian Infantry Battalion guarded the 
line of communications, the 2/25th Aus- 
tralian Infantry Battalion advanced 
down the road and part of the 2/ 2d 
Australian Pioneer Battalion moved 
down the north bank of the river. 

When a small group of Japanese of- 
fered resistance to the advance at Jen- 
sen's Plantation, toward the lower end 
of the valley, the 2/25th Battalion drove 
it back and on 14 September captured 
Heath's Plantation farther on. The 
2/33d Australian Infantry Battalion 
then took over and pushed on toward 
Lae. By now the Australians had come 
within range of Japanese 75-mm. guns 
and found the going harder. But an as- 
sault the next day cleared Edward's 
Plantation and enemy resistance ended. 

The advance elements of the 25th Bri- 
gade entered Lae from the west the next 
morning, 16 September. In the after- 
noon the 24th Brigade, which had ad- 
vanced from the east and captured Mala- 
hang Airdrome on 15 September, pushed 
into Lae and made contact with the 
25th Brigade. Lae had fallen easily and 
speedily. The Japanese had vanished. 



The Japanese Evacuation 

Throughout July and August the Sala- 
maua Japanese were reinforced at Lae's 
expense, but were continually forced 
back. On 24 August General Nakano, 
reflecting the importance which his su- 
perior had attached to Salamaua, ad- 
dressed his troops thus: "Holding Sala- 
maua is the Division's responsibility. 
This position is our last defense line, 
and we will withdraw no further. If we 
are unable to hold, we will die fighting. 
I will burn our Divisional flag and even 
the patients will rise to fight in close 
combat. No one will be taken a pris- 
oner." 20 

Imperial Headquarters, however, did 
not order a suicidal last stand. Nakano 
was ordered to hold out as long as pos- 
sible, but to withdraw if he could not 
hold Salamaua. The Australian landing 
between Lae and Finschhafen and the 
503d's seizure of Nadzab, coupled with 
Allied air and PT boat activity in the 
Huon Gulf and the straits, caused Gen- 
eral Adachi on 8 September to order 
Nakano to abandon Salamaua and pull 
back to Lae. Nakano's hospital patients 
and artillery had already been sent to 
Lae, and on 1 1 September withdrawal 
of the main body began. 

Meanwhile, after considerable discus- 
sion Imperial Headquarters, Imamura, 
and Adachi abandoned their plans to 
take Bena Bena and Mount Hagen. 
Adachi saw that the Allied operations at 
Salamaua, Nadzab, and Lae threatened 
to cut off the 51st Division. He now de- 
cided that he would have to withdraw 
from Lae, but determined to hold the 
Finisterre Range, the Ramu Valley, and 

* 18th Army Operations, II, Japanese Monogr No. 
42 (OCMH), 37. 

Finschhafen. Therefore he ordered 
Nakano and Shoge to withdraw overland 
from Lae to the north coast of the Huon 
Peninsula, and directed the 20th Di- 
vision to move from Madang to Finsch- 
hafen and to dispatch a regiment to the 
Ramu Valley to assist the 51st. 

Thus the Allied troops pushing to- 
ward both Lae and Salamaua in early 
September met only delaying forces. The 
Salamaua garrison had assembled at Lae 
by the 14th, two days after the first 
echelon of the Lae garrison had started 
north. Another echelon left that day, and 
the last slipped out on the 15th. The day 
before, General Vasey had learned from 
a captured document and from inter- 
rogation of a prisoner that the Japanese 
were leaving Lae. He dispatched troops 
northward to reinforce the 2/4th Aus- 
tralian Independent Company, which 
was operating in the wilds north of Lae, 
but the Japanese eluded their pursuers. 
It was a band of retreating enemy that 
the 3d Battalion, 503d, encountered at 
Yalu, and when Australian forces rushed 
there the Japanese hastily altered their 
route to avoid interception. 

Once out of Lae, the 51st Division and 
the Lae naval garrison executed one of 
the difficult overland marches that were 
to characterize so many future Japanese 
operations in New Guinea. There was 
little fighting, but Australian patrols 
harried the retreat. The Japanese moved 
north out of Lae and avoided Nadzab 
and the obvious Markham-Ramu trough 
that Adachi had originally planned to 
use for the withdrawal. They moved in 
a generally north-northeasterly direc- 
tion, crossed the Busu River by means 
of a rough-hewn bridge on 20-22 Sep- 
tember, and skirted the west ends of 



the Rawlinson Range and Cromwell 
Mountains in the vicinity of Mount Sala- 
waket about 25 September. 

They had started with food for ten 
days, but this was exhausted by the time 
they reached Salawaket. Thereafter they 
lived by looting native gardens and by 
eating roots and grasses. Dysentery and 
malaria made their appearance, but as 
there were plenty of suppressive drugs 
the malaria rate was low. 

The 51st Division had already aban- 
doned most of its heavy equipment be- 
fore the retreat. Along the way mountain 
artillerymen, unable to drag their guns 
over the precipitous slopes, were forced 
to abandon them. Many soldiers threw 
away their rifles. This was in strong con- 
trast to the behavior of the 1st Battalion, 
20th Division, which had reinforced the 
51st Division at Salamaua. The com- 
mander, a Major Shintani, had threat- 
ened death to any soldier who abandoned 
his arms. Shintani died on the road, but 
his battalion rigorously adhered to his 
orders. Each soldier who completed the 
march carried his rifle and his helmet. 

By mid-October the troops reached the 
north coast of the Huon Peninsula. The 
Army troops went to Kiari, naval per- 
sonnel to nearby Sio. Slightly over 9,000 
men had left Lae; 600 were march 
casualties. Nearly 5,000 soldiers arrived 
at Kiari, and some 1,500 sailors went to 
Lio. Many others were taken to the hos- 
pital at Madang. The defense of Lae- 
Salamaua and the subsequent retreat 
cost almost 2,600 lives. 

Strategic Reconsiderations 

The Japanese Pull Back 

The fall of Lae and Salamaua, coming 
hard on the heels of defeat in the central 

Solomons, had a profound effect upon 
Japanese strategic plans, an effect that 
went far beyond the immediate impor- 
tance of Lae and Salamaua. Although 
the twin losses of Guadalcanal and Buna 
were severe, Imperial General Head- 
quarters had not regarded these as irre- 
trievable. It had continued to prepare 
plans for offensives in the Southeast 
Area. 21 Now the war leaders in Tokyo 
reassessed the situation and determined 
on a drastic retrenchment. 

The fall of the central Solomons and 
of Lae-Salamaua closely followed the 
loss of Attu and the evacuation of Kiska 
in the Aleutians, and came at a time 
when Imperial Headquarters enter- 
tained well-justified fears about the 
opening of an Allied offensive through 
the Central Pacific. 22 The Japanese in 
September decided that they were over- 
extended. They determined to withdraw 
their perimeter in order to set up a de- 
fense line that would hold back the 
Allies while they themselves marshaled 
their strength for decisive battle. This 
perimeter would be strongly manned 
and fortified. It was hoped that the de- 
fensive preparations behind it would be 
completed by the early part of 1944. 

So Imperial Headquarters drew its 
main perimeter line from western New 
Guinea through the Carolines to the 
Marianas. This was "the absolute na- 
tional defense line to be held by all 

"This subsection is based upon: 8th Area Army 
Operations, Japanese Monogr No. no (OCMH), 
pp. 80-87; an d 18th Army Operations, II, Japanese 
Monogr No. 42 (OCMH), 19-22, 48—53, 58-64, 84- 
86, 151-54. 

a The offensive began in November 1943. See 
Philip A. Crowl and Edmund G. Love, The Seizure 
of the Gilberts and Marshalls, UNITED STATES 
ARMY IN WORLD WAR II (Washington, 1955). 



means." 23 The Southeast Area, includ- 
ing Rabaul, once the focus of such great 
but elusive hopes for victory, was now 
on the outpost line. 

But the war was far from over for 
MacArthur's and Halsey's troops. Gen- 
eral Imamura and Admiral Kusaka were 
no longer counted on to win decisively, 
but they were ordered to hold out as 
long as possible, and so delay the Allied 
advance. To strengthen the Southeast 
Area, Imperial Headquarters in Septem- 
ber ordered the lyth Division from 
Shanghai to Rabaul "to reinforce the 
troops manning the forward wall." 24 

Imamura and Kusaka determined to 
hold Bougainville, whose defenses they 
had been trying to build up during the 
long fight on New Georgia, to develop 
and strengthen Madang and Wewak, to 
develop the transport system connecting 
the main bases of the Southeast Area, and 
to hold Dampier and Vitiaz Straits. Con- 
trol of these straits had been essential to 
nearly all Japanese movements to New 
Guinea and, as before, the Japanese were 
resolved to hold them in order to block 
any A llied westward advance. (See Map 


To this end Imamura, who kept the 
38th Division under his control to de- 
fend Rabaul, had previously dispatched 
the reinforced 65th Brigade to Tuluvu 
on the north coast of Cape Gloucester 
with orders to develop a shipping point 
there and to maintain the airfield. On 5 
September he sent Maj. Gen. Iwao Mat- 

13 8th Area Army Operations, Japanese MonogT 
No. no (OCMH), p. 87. 

"Japanese Operations in the Southwest Pacific 
Area (Vol. II of MacArthur hist), Ch. VII, p. 26, 

suda to Tuluvu to take command of the 
65th Brigade, some elements of the 51st 
Division, and the 4th Shipping Group. 
To Matsuda's responsibility for han- 
dling shipping he added that of defend- 
ing the coasts of western New Britain. 

On the New Guinea side of the straits, 
the Japanese regarded Finschhafen as the 
key defensive position. Possessed of two 
good harbors— Finschhafen itself and 
Langemak Bay— and a small airfield, it 
had long been used as a barge staging 
point. In early August Adachi had been 
concerned about a possible attack against 
Finschhafen, but he did not have enough 
troops to strengthen its small garrison 
substantially while the 41st Division was 
defending Wewak, the 51st Division was 
defending the Lae-Salamaua area, and 
the 20th Division was working on the 
Madang-Lae road. He did, however, 
send the 80th Infantry and one battalion 
of the 21st Field Artillery Regiment of 
the 20th Division from Madang to 
Finschhafen. By the end of August Maj. 
Gen. Eizo Yamada, commanding the 1st 
Shipping Group and the combat troops 
at Finschhafen, had about one thousand 

When the 9th Australian Division 
landed east of Lae on 5 September, 
Adachi foresaw the danger, to Finsch- 
hafen. He suspended construction of the 
Madang-Lae road, which was now a 
twenty-foot- wide all-weather road run- 
ning along the coast from Madang to 
Bogadjim, thence over the Finisterre 
Range at a defile named Kankirei and 
into the Ramu Valley to a point ten 
miles north of Dumpu. This decision 
freed the 20th Division for combat duty. 
Adachi ordered a small force of the 20th 



Division, under Maj. Gen. Masutaro 
Nakai, the divisional infantry com- 
mander, to advance to Kaiapit, which is 
on the uplands near the sources of the 
Markham and Ramu Rivers. The move 
was intended to keep the Allies from ad- 
vancing through the Ramu Valley, over 
Kankirei to the coast, and on against 
Madang and Wewak, and was also to 
help cover the retreat of the 51st Di- 
vision from Lae up the trough to Ma- 
dang. When Adachi decided not to use 
the Markham and Ramu Valleys for the 
retreat he ordered Nakai north to hold 
the Kankirei defile. 

Adachi ordered the main body of the 
20th Division, commanded by I_t. Gen. 
Shigeru Kitagiri, to march to Finsch- 
hafen. The division departed Bogadjim 
on 10 September on its march of nearly 
two hundred miles, but was still far from 
its destination when the Allies struck the 
next blow. 

Allied Decisions 

General Mac Arthur's Elkton III 
called for the capture of Finschhafen as 
a step toward gaining control of Vitiaz 
and Dampier Straits. The plan had set 
the tentative date for the move against 
Finschhafen at six weeks after the in- 
vasion of Lae. At least two factors, how- 
ever, impelled a speed-up in the time- 
table. The first was the quick fall of Lae 
and Salamaua after the landing on 4 
September. The second was the 20th Di- 
vision's move toward Finschhafen. But 
before orders could be sent out for the 
capture of Finschhafen, it was necessary 
to consider this operation in relation to 
the larger problems involved in captur- 
ing Madang, an operation considered 
necessary to protect the left flank during 

the seizure of Cape Gloucester. Seizure 
of Finschhafen, Madang, and Cape 
Gloucester would of course give physical 
control of both sides of the straits to the 
Allies. 25 

Capture of Madang was bound to be 
a large operation. Allied intelligence 
estimated that in late August a total of 
55,000 Japanese held the regions be- 
tween Lae and Wewak. At this time 
General Blarney, in a letter to Mac- 
Arthur, held that the Japanese would 
exert every effort to defend the Mark- 
ham and Ramu Valleys, Bogadjim (near 
the defile into the Ramu), Lae, Sala- 
maua, and Finschhafen. Capture of 
Madang, which had been assigned to his 
New Guinea Force, would require as 
preliminary conditions complete air and 
naval superiority, support by the VII 
Amphibious Force, physical possession of 
Lae, the Markham Valley, Salamaua, and 
Finschhafen, and the neutralization of 
the Japanese in western New Britain. 

Blarney set forth three steps to be 
followed after the capture of Lae. The 
first was the capture of Finschhafen by 
a seaborne assault. Blarney recommended 
as the second step seizure of an inter- 
mediate objective between Finschhafen 
and Bogadjim, because 256 miles of 
water separated Lae from Bogadjim, 

!s See above pp. 27, 190 and Chart 2. This subsec- 
tion is based on ALF, Rpt on New Guinea Opns: 4 
Sep 43-26 Apr 44; Ltr, Blarney to MacArthur, 31 Aug 
43, no sub, in GHQ SWPA G-3 Jnl, 31 Aug 43; 
Memo, Chamberlin for CofS GHQ SWPA, 3 Sep 
43, sub: Comment on Ltr From Cmdr ALF, 31 Aug 
43, Opns for Capture of Madang, same file; Cham- 
berlin's Memo for File, 3 Sep 43, same file; and 
GHQ SWPA OI 34A2, 15 Sep 43, in GHQ SWPA 
G-3 Jnl, 14 Jul 43; GHQ SWPA OI 34/14, 17 Sep 
43, same file. 



178 separated Finschhafen and Bogad- 
jim, and these were long distances to 
travel with a flank exposed. The third 
and final step would be the capture of 
Madang by a combination of airborne 
invasion and amphibious assault coupled 
with pressure from troops advancing 
northwestward out of the Ramu Valley. 
To avoid exposing the right flank, he 
strongly urged capturing Cape Glouces- 
ter (which had been assigned to the 
Alamo Force) before taking Madang. 
This would be feasible, he argued, be- 
cause Madang was so much farther from 
Finschhafen than was Cape Gloucester. 

These proposals received close study 
at the advanced echelon of GHQ, which 
had moved to Port Moresby during the 
planning for Lae and the Markham 
Valley. General Chamberlin looked on 
them as generally sound. Regarding 
Blarney's concern over control of Cape 
Gloucester as well as the coasts of the 
Huon Peninsula, however, he pointed 
out to General Sutherland that "G-g be- 
lieves that a physical occupation of areas 
has little bearing on the control of Vitiaz 
Strait but considers that airfields stra- 
tegically placed which cover the water 
areas north of Vitiaz Strait are the con- 
trolling considerations." 26 

As to the intermediate objective be- 
tween Finschhafen and Bogadjim, which 
Chamberlin placed at Saidor (with a 
harbor and prewar airfield) on the north 

"Memo, Chamberlin for CofS GHQ SWPA, 3 
Sep 43, sub: Comment on Ltr From Comdr ALF, 
31 Aug 43, Opns for Capture of Madang, in GHQ 
SWPA G— 3 Jnl, 31 Aug 43. Japanese sources often 
use "Dampier Strait" to mean both Vitiaz and 
Dam pier Straits; Allied sources often use "Vitiaz 
Strait" for both. 

coast of the Huon Peninsula, he felt that 
little would be gained by seizing it as 
well as Madang. On the other hand it 
appeared that Saidor might prove a 
satisfactory substitute for Madang. 

Timing of operations would be 
tricky, largely because the VII Am- 
phibious Force lacked enough ships to 
conduct two operations at once. It would 
be committed to operations on the Huon 
Peninsula until mid-November. There- 
fore the Cape Gloucester invasion could 
not take place until about 1 December, 
but the attack against the north coast of 
the Huon Peninsula would also have to 
be launched about the same time if the 
New Britain offensive was to be pro- 
tected effectively. For these reasons 
Chamberlin recommended deferring the 
decision on whether to move to New 
Britain before or after invading the 
north coast of the Huon Peninsula. For 
this latter operation, he proposed two 
alternatives: seizure of a prewar airfield 
at Dumpu in the Ramu Valley without 
operating on the coast at all, or seizure 
of the Saidor airfield without operating 
in the Ramu Valley. 

The questions were threshed out at a 
conference at Port Moresby on 3 Sep- 
tember. MacArthur, Sutherland, Cham- 
berlin, Kenney, Whitehead, Blarney, 
Carpender, and others attended. Blarney 
spoke strongly in favor of his recom- 
mendations. Kenney urged a deep pene- 
tration of the Ramu Valley all the way 
to Hansa Bay, which lies between Ma- 
dang and Wewak. After Hansa Bay, he 
recommended, the advance could turn 
southward in co-ordination with the 
Cape Gloucester attack. Admiral Car- 
pender wanted an operation somewhere 



between Madang and Saidor to precede 
Cape Gloucester. He received some sup- 
port in his view from MacArthur, who 
asserted the necessity for seizing an area 
between Finschhafen and Madang be- 
fore capturing Cape Gloucester, so as to 
assure the safe movement of supplies to 
support the latter operation. After a 
good deal of discussion, opinion crystal- 
lized in favor of covering the move to 
Gloucester by seizing the line Dumpu- 
Saidor. Dumpu would be seized at once 
by airborne and overland advances, and 
would then be used to cover simultane- 
ous moves against Saidor and Gloucester. 
These moves, Chamberlin estimated on 
3 September, would take place about 1 
November at the earliest, but i Decem- 
ber was more probable. 

Thus it was that on 15 September 
MacArthur ordered Blarney's New 
Guinea Force, supported by Kenney's 
forces, to seize Kaiapit at the head of 
the Markham Valley and Dumpu about 
thirty miles south of Bogadjim. Two 
days later he ordered the New Guinea 
Force, with naval support, to capture 
Finschhafen. It would serve as an ad- 
vanced air base, and Allied Naval Forces, 
basing light naval craft there, would use 
it to cut off the Japanese from Cape 
Gloucester and Saidor. The attack on 
Madang was postponed. 

Advance Through the Ramu Valley 

With his forces converging on Lae 
from east and west General Blarney com- 
pleted plans for Kaiapit and Dumpu. 
Tactically the initial phases of the task 
appeared fairly simple; patrols had re- 
ported the area between Nadzab and 
the Leron River, a tributary of the 

Markham, to be free of the enemy. 27 
Logistics would present the greatest dif- 
ficulty. No overland line of communica- 
tions existed, and until roads were built 
all supplies for the advancing troops 
would have to be flown in. This fact 
limited the attacking force to one di- 
vision (the 7th) of but two brigade 

The 2/6th Australian Independent 
Company began the drive in September 
when Kenney's transport planes landed 
it on a prewar airstrip in the Markham 
Valley some thirty miles northwest of 
Nadzab near the Leron River. The 2/6th 
then made its way eight miles up the 
river to Kaiapit, after a sharp encounter 
on 19 September, captured the village 
from a small group of Japanese, and 
held it against their repeated counter- 
attacks. Two days later the Kaiapit strip 
saw the arrival, after a flight up from 
Nadzab, of the 21st and 25th Brigade 
Groups of General Vasey's 7 th Aus- 
tralian Division. 

At the month's end the 21st Brigade, 
followed by the 25th, left Kaiapit and 
entered the Ramu Valley. By 6 October 
the 21st was in possession of Dumpu, 
where 7th Division headquarters was 
established. The great Markham-Ramu 
trough had fallen with an ease that the 
Allies had not expected, an ease brought 
about by the hasty Japanese decision not 
to retreat through the trough. 

Behind the lines engineers set to work 
building a truck highway from Lae to 
Nadzab along the prewar road, but rain 
fell during forty-six of the final sixty 

21 This section is based on ALF, Rpt on New 
Guinea Opns: 4 Sep 43-26 Apr 44; Craven and 
Cate, The Pacific: Guadalcanal to Saipan, pp. 190— 




22 3»p»«mb«r-ZO October 1943 



Form lint inlttnol *Q0 fnt 

o i e s « MILES 

I 1 — '-I r 1 1 ' 1 



>— hrSsltietz-a — ^"^aK 


MAP 14 

days of the project and it was December 
before the task was finally finished and 
large amounts of supplies could be sent 
to Nadzab. Nadzab and the other sites in 
the Markham and Ramu Valleys re- 
ceived all their supplies and equipment 
by airlift during the period the road was 
under construction. 

By the end of December Allied Air 
Forces possessed three first-class air bases 
in full-scale operation in the Markham 
and Ramu Valleys: one at Nadzab, one 
at Lae, and one at the juncture of the 
Gusap and Ramu Rivers. The last site 
was selected in preference to Kaiapit, 
which proved too swampy and malarious 
for extensive development. Dumpu 
served as a staging field for fighter planes. 

After establishing strong positions at 
Dumpu, the 7th Australian Division 
continued its part in seizing the Huon 
Peninsula. Marching north-northwest 
from Dumpu, it attacked Nakai's posi- 
tions in the defiles of the Finisterres. The 
defiles were secured in February after 
almost three months of the most arduous 
kind of fighting. Nakai retreated toward 
Madang while Vasey's division broke out 
to the coast east of Madang. 

The Coastal Advance 


Vasey's operations through the Ramu 
Valley were co-ordinated with those of 
Wootten's gth Australian Division, 
which was operating on the coasts of the 
Huon Peninsula in a series of operations 
that began with Finschhafen. Before 
leaving Milne Bay for Lae, Wootten had 
been alerted to the possibility that he 
might have to send a brigade to Finsch- 
hafen. (Map 14) Thus GHQ's decision 
on 17 September to invade Finschhafen 



at once was no surprise to the veteran 
Australian commander. 28 Admiral Bar- 
bey had just time enough, and no more, 
to assemble 8 LST's, 16 LCI's, 10 de- 
stroyers, and 4 APD's for the invasion on 
22 September, but "Uncle Dan" was now 
an old amphibious hand and he met the 
deadline. The LST's loaded at Buna, and 
the whole task group assembled in the 
harbor at Lae on 2 1 September. General 
Wootten meanwhile had selected the 
20th Infantry Brigade Group of his di- 
vision to make the landing, and had 
ordered the 22d Infantry Battalion to 
advance east along the coast to threaten 
Langemak Bay, just south of Finsch- 
hafen. Elements of the 2d Engineer Spe- 
cial Brigade had been attached to Woot- 
ten, and these units also made ready. No 
close air support was planned for the in- 
vasion, but in the days preceding 22 
September B-24's and B-25's bombed 
the Gasmata airfield on the south coast 
of New Britain. Daytime A-20's and 
B-25's struck at Japanese lines of com- 
munication to Finschhafen, and PT's 
took over the work at night. 

Troops of the 20th Brigade boarded 
their convoy on the afternoon of 2 1 Sep- 
tember. The force included, besides the 
Australians and Barbey's American 

" This section is based on Craven and Cate, The 
Pacific: Guadalcanal to Saipan, pp. 187-89; Mori- 
son, Breaking the Bismarcks Barrier, pp. 269—74; 
Off of Chief Engr, GHQ AFPAC, Engineers in 
Theater Operations, p. 112, and Critique, pp. 106— 
09; ALF, Rpt on New Guinea Opns: 4 Sep 43—26 
Apr 44; 2d ESB Rpt; GHQ SWPA Check Sheet, 
Chief Engr SWPA to CINC, CofS, and G-3 SWPA, 
23 Oct 43 [a rpt of 2d ESB action], in GHQ SWPA 
G— 3 Jnl, 23 Oct 43; Japanese Operations in the 
Southwest Pacific Area (Vol. II of MacArthur hist), 
Ch. VII, OCMH; 8th Area Army Operations, Japa- 
nese Monogr No. 110 (OCMH), pp. 82-83; •Sth 
Army Operations, III, Japanese Monogr No. 43 
(OCMH), 84-117. 

sailors, one boat company, half the shore 
battalion of the 532d Engineer Boat and 
Shore Regiment, and medical and signal 
troops, or 575 men, 10 LCM's, and 15 
LCVP's. The 22d Battalion marched out 
of Lae en route to Langemak Bay on the 
21st, and the same day the amphibious 
force sailed for Finschhafen, eighty-two 
miles distant. 

The beach selected for the landing, 
designated Scarlet, lay six miles north 
of Finschhafen at the mouth of the Song 
River. It was nine hundred yards long 
(north to south), thirty feet wide, and 
was marked by coral headlands to the 
north and south. 

Destroyers bombarded Scarlet Beach 
on the morning of 22 September, and 
during darkness, at 0445, the first Aus- 
tralian assault wave touched down. 
Coxswains had difficulty finding the right 
beach in the dark with the result that 
most landing craft carrying the first two 
waves lost direction and landed in a 
small cove south of Scarlet Beach. First 
light aided the LCI's carrying the third 
wave; they landed at the right place. The 
waves that landed at the cove met some 
scattered but ineffective fire from enemy 
posts in the fringe of the jungle. The 
third wave met better organized opposi- 
tion from log-and-earth pillboxes, but 
by 0930 all resistance had been over- 
come, all troops and supplies were 
ashore, and the landing craft retracted. 
The Japanese survivors retired to rising 
ground about a half mile inland and 
some sharp fighting ensued before the 
2/ 17th Battalion was in complete pos- 
session of the beachhead. The 2/13 Bat- 
talion meanwhile swung left (south) to- 
ward the village of Heldsbach, which was 
just north of the Finschhafen airstrip. 



General Yamada had posted only a 
small part of his force at Scarlet Beach. 
He was keeping the rest of his 4,000-man 
command at Hanisch Harbor on the 
south coast of the peninsula and on 
Satelberg, a 3,240-foot peak which was 
about six miles west of Scarlet Beach, 
dominated the entire coastal region, and 
overlooked both Finschhafen and Lange- 
mak Bay. When General Adachi re- 
ceived news of the Allied landing he 
ordered Yamada to concentrate his force 
at Satelberg and attack at once. This at- 
tack was designed to hold or destroy the 
Australians pending the arrival of Gen- 
eral Kitagiri's 20th Division. By 21 Sep- 
tember the 20th Division, advancing 
overland and hauling its heavy materiel 
on barges, had reached Gali, one hun- 
dred miles from Finschhafen; it expected 
to arrive at Finschhafen on 10 October. 
Adachi ordered Kitagiri to hurry. 

Admiral Barbey's retiring ships offered 
a tempting target to Japanese airmen, 
but the yth Air Division, under orders 
to cover a Wewak-bound convoy, hesi- 
tated to leave it unprotected. The 4th 
Air Army headquarters ended this inde- 
cision by ordering the yth out against 
Barbey, but bad weather over central 
New Guinea kept the Army planes on 
the ground. Those of the naval nth Air 
Fleet at Rabaul went up and fiercely at- 
tacked the amphibious force on 22 and 
24 September. But the vigilant destroyer 
Reid had given warning and Allied fight- 
ers, the ships' own antiaircraft, and "good 
luck in addition to good ship maneuver- 
ing" kept the ships from harm. 29 

At the beachhead the American en- 
gineers built roads and dumps and un- 

"Ltr, Adm Barbey to Gen Smith, Chief of Mil 
Hist, 20 Nov 53, no sub, OCMH. 

loaded naval craft. The larger engineer 
craft carried additional supplies from 
Lae to Scarlet Beach, while the LCVP's 
hauled supplies at night from Scarlet 
Beach to the Australians who were push- 
ing south toward Finschhafen. 

The 20th Brigade continued its move 
toward Finschhafen on the 23d. It cap- 
tured Heldsbach, the airfield, and part 
of the shore of the harbor before meeting 
stiff resistance at the Bumi River, where 
three hundred enemy sailors and one 
company of the 2d Battalion, 238th In- 
fantry, defended the south bank. Two 
companies of the 2/ 15th Battalion moved 
inland (right) to outflank the enemy, 
and the next morning the Australians 
forced their way over the river in the 
face of stalwart resistance. The brigade 
commander, who was becoming increas- 
ingly aware of the Japanese concentra- 
tion at Satelberg, asked Wootten for one 
more battalion with which to hold Scar- 
let Beach while he concentrated his bri- 
gade against Finschhafen. Wootten as- 
sented. The 2/43d Battalion landed at 
Scarlet Beach on the night of 29-30 Sep- 
tember to relieve the 2/ 17th, and the 
latter moved out at once for Finsch- 
hafen. Following air and artillery bom- 
bardment, the three Australian battal- 
ions—the 2/1 3th, 2/ 15th, and 2/ 17th— 
attacked on 1 October, fought all day, 
and overwhelmed the defenders. The 
next morning they occupied the village 
and harbor of Finschhafen and made 
contact south of Langemak Bay with 
patrols of the 2 2d Battalion, which had 
advanced overland from Lae. 

The Counterattack 

To gain complete control of the New 
Guinea side of Vitiaz Strait, Generals 



MacArthur and Blarney had ordered 
that the capture of Finschhafen be fol- 
lowed by an advance along the coast to 
Sio, fifty land miles distant. But the ad- 
vance could not be undertaken until 
the Japanese were driven from their 
dominating positions at Satelberg and 
on Wareo spur, a lower spur which lay 
north o£ the Song River from Satelberg. 

On 26 September Yamada had 
launched an unsuccessful attack with the 
80th Infantry against the Australian 
beachhead. After Finschhafen fell on 2 
October, the 20th Brigade moved back 
to Scarlet Beach in preparation for an 
assault against Satelberg. Two battalions 
attacked but met stout resistance. 

When General Wootten's headquar- 
ters and the 24th Brigade arrived, Woot- 
ten decided that all signs indicated the 
Japanese would counterattack immedi- 
ately, before he could complete his 
preparations for advancing to Sio. He 
decided to go on the defensive for the 
time being. 

Meanwhile the 20th Division was on 
its way; advance elements totaling 2,354 
men had reached Sio by 30 September. 
General Kitagiri decided to advance by 
an inland route rather than use the 
coastal track to Satelberg. Like so many 
other Japanese generals in similar cir- 
cumstances during World War II, he 
decided not to concentrate all his forces 
before attacking but ordered his units 
to attack the Australians upon arriving. 
Japanese tactical doctrine warns of the 
dangers of such piecemeal commitment 
but Japanese generals frequently aided 
the Allied cause by putting aside their 
doctrine in favor of pell-mell, piecemeal 

For his main attack against Scarlet 

Beach Kitagiri decided to drive east- 
ward from Satelberg with most of his 
forces while a small detachment aboard 
four landing craft attempted an am- 
phibious assault. But his division was 
no better at safeguarding important 
documents than was any other Japanese 
unit. On 15 October General Wootten 
received a captured Japanese order which 
warned him to expect a two-regiment at- 
tack from Satelberg, coupled with a sea- 
borne assault. The Australians made 

Next day the 9th Division, though suf- 
fering some local reverses, repulsed the 
2oth's attack from Satelberg. At 0300, 17 
October, Japanese planes bombed the 
Allies, whereupon 155 men of the 10th 
Company, jgth Infantry, attempted to 
land from their four craft. Two barges 
were sunk, one departed in haste, and 
the other reached shore in the vicinity 
of a .50-caliber machine gun position 
manned by Pvt. Nathan Van Noy, Jr., of 
the 5 3 2d Engineer Boat and Shore Regi- 
ment, and one other American engineer. 
As the enemy soldiers disembarked they 
hurled grenades, one of which wounded 
Van Noy before he opened fire. But Van 
Noy held his fire until the Japanese were 
visible, then opened up and killed about 
thirty of them. He died of his wounds, 
and for his gallant devotion was awarded 
the Medal of Honor. 30 Though the Jap- 
anese claim that the few men who 
reached the shore wrought great dam- 
age, in actuality they were all quickly 

Later in the morning came another 
major attack from Satelberg. Wootten, 

! "See The Medal of Honor 0} the United States 
Army (Washington, 1948), pp. 283-84. Van Noy's 
loader, who was wounded, received the Silver Star. 



who had no reserve brigade, asked for 
the 26th and Barbey's ships transported 
it to Scarlet Beach on 20 October. The 
Japanese attacks continued through 25 
October, but all failed. As his food sup- 
plies were exhausted, Kitagiri suspended 
the attacks and regrouped for another 
try. The Australians, losing 49 dead in 
these actions, reported killing 679 of 
the enemy. 

General Adachi, who had often been 
in and out of Salamaua during the fight- 
ing there, traveled from Madang via 
Kiari and Sio to Satelberg. He arrived 
on 31 October, and stayed for four days. 
During this period Kitagiri made some 
hopeful estimates on the success of 
future, more gradual offensives. 

Satelberg to Sio 

But Wootten was now ready to as- 
sume the offensive. By 17 November one 
more brigade, the 4th, had arrived to 
hold the beachhead while the three in- 
fantry brigades of the 9th Division at- 
tacked. Meanwhile work on the airstrip 
and advanced naval base at Finschhafen 
had gone forward so quickly that PT 
boats from Finschhafen were now harry- 
ing enemy sea communications at night 
in consort with PBY's ("Black Cats"). 
With the support of tanks and artillery, 
and rocket-equipped LCVP's lying off- 
shore, the 9th Division fought a major 

action starting on 17 November. By 8 
December it had captured Satelberg and 
Wareo spur and was ready to push up 
the coast to Sio, whence the 20th Di- 
vision was retreating on orders from 
General Imamura himself. 

Wootten's men advanced slowly but 
steadily against the retreating enemy, 
supported all the while by the 2d En- 
gineer Special Brigade craft. 31 The Aus- 
tralians found many sick, wounded, and 
dead Japanese who had fallen by the 
way as the weakened 20th Division, 
which numbered 12,526 men on 10 Sep- 
tember and only 6,949 men by Decem- 
ber, laboriously marched along. On 15 
January 1944 the 9th Division entered 
Sio, on the north coast of the Huon 

Fighting on the peninsula was not yet 
over, but the main strategic objectives— 
the airfield sites and the coast of Vitiaz 
Strait— were now in Allied hands. When 
the Lae-Nadzab road and the airfields 
were completed, the Allies could control 
the air over the straits and bring a 
heavier weight of metal to bear on Japa- 
nese bases to the north and to the west. 

11 In the Nassau Bay— Lae-Finschhafen operation 
the 2d Engineer Special Brigade lost twenty-one 
dead, ninety-four wounded, and sixty evacuated 
sick. On the pursuit to Sio four LCVP's were lost 
to enemy action, four more to surf and reefs. 


The Invasion of Bougainville 

While MacArthur's and Halsey's 
troops were gaining the Trobriands, the 
Markham Valley, the Huon Peninsula, 
and the New Georgia group for the 
Allied cause, the Joint Chiefs of Staff 
and their subordinate committees in 
Washington had been making a series 
of decisions affecting the course of the 
war in the Pacific. These decisions re- 
lated not so much to Cartwheel itself 
as to General MacArthur's desire to 
make the main effort in the Pacific along 
the north coast of New Guinea into the 
Philippines. But, since they called for 
troops to support the offensives in Ad- 
miral Nimitz' Central Pacific Area, they 
had an immediate impact upon Cart- 
wheel, especially on the Bougainville 
invasion (Operation B of Elkton III) 
and on MacArthur's plans to seize 
Rabaul and Kavieng after Cartwheel. 

The Decision To Bypass Rabaul 

Once the Combined Chiefs at Casa- 
blanca had approved an advance through 
the Central Pacific, the Joint Chiefs put 
their subordinates to work preparing a 
general strategic plan for the defeat of 
Japan. An outline plan was submitted at 
the meeting of the Combined Chiefs in 
Washington, 12-15 May 1943. The Com- 

bined Chiefs approved the plan as a basis 
for further study. 1 

The plan, which governed in a general 
way the operations of Nimitz* and Mac- 
Arthur's forces until the end of the war, 
aimed at securing the unconditional sur- 
render of Japan by air and naval block- 
ade of the Japanese homeland, by air 
bombardment, and, if necessary, by in- 
vasion. The American leaders agreed 
that naval control of the western Pacific 
might bring about surrender without 
invasion, and even without air bombard- 
ment. But if air bombardment, invasion, 
or both proved necessary, air and naval 
bases in the western Pacific would be re- 
quired. Therefore, the United States 
forces were to fight their way westward 
across the Pacific along two axes of ad- 
vance: a main effort through the Cen- 
tral Pacific and a subsidiary effort 
through the South and Southwest Pacific 
Areas. 2 

The Washington commanders and 

1 Cline, Washington Command Post, pp. 219-22. 

2 In point of fact the terms "main effort" and 
"subsidiary effort," though used constantly, bore so 
little relation to the number of troops, aircraft, and 
ships engaged as to be almost without meaning. In 
general, the Central Pacific had the preponderance 
of fleet (included carrier-based air) strength; Mac- 
Arthur had the greater number of divisions and 
land-based aircraft. 



planners preferred the Central Pacific 
route for the main effort because it was 
shorter and more healthful than the 
South-Southwest Pacific route; it would 
require fewer ships, troops, and supplies; 
success would cut off Japan from her 
overseas empire; destruction of the Japa- 
nese fleet, which would probably come 
out fighting to oppose the advance, would 
enable naval forces to strike directly at 
Japan; and it would outflank and cut off 
the Japanese in the Southeast Area. The 
main effort should not be made through 
the South and Southwest Pacific Areas, 
it was argued, because a drive from New 
Guinea to the Philippines would be a 
frontal assault against large islands with 
positions closely arranged in depth for 
mutual support. The Central Pacific 
route, in contrast, permitted the con- 
tinuously expanding U.S. Pacific Fleet 
to strike at small, vulnerable positions 
too widely separated for mutual support. 

The Joint Chiefs decided on the two 
axes, rather than the Central Pacific 
alone, because the Japanese conquests 
in the first phase of the war had com- 
pelled the establishment of compara- 
tively large Allied forces in the South and 
Southwest Pacific Areas; to shift all these 
to the Central Pacific would take too 
much time and too many ships, and 
would probably intensify the already 
strong and almost open disagreement be- 
tween MacArthur and King over Pacific 
strategy. Further, the Joint Chiefs hoped 
to use the oilfields on the Vogelkop 
Peninsula. 3 Twin drives, co-ordinated 
and timed for mutual support, would 

'This hope came to nothing. Robert Ross Smith, 
The Approach to the Philippines, UNITED 
ington, 1953), pp. 425-28. 

give the U.S. forces great strategic ad- 
vantages, for the Japanese would never 
know where the next blow would fall. 4 

At Washington in May the Combined 
Chiefs, as they had at Casablanca, ap- 
proved plans for seizure of the Gilbert 
and Marshall Islands as the opening 
phase of the Central Pacific advance. 
They also approved the existing plans 
for Cartwheel, which the Joint Chiefs 
estimated would be ended by April 1944. 

Next month, the Joint Chiefs, con- 
cerned with the problem of co-ordinating 
Nimitz' and MacArthur's operations, 
asked MacArthur for specific informa- 
tion on organization of forces and dates 
for future operations and informed him 
that they were planning to start the Cen- 
tral Pacific drive in mid-November. They 
planned to use the 1st and 2d Marine 
Divisions, then in the Southwest and 
South Pacific Areas, respectively, all the 
South Pacific's assault transports and 
cargo ships (APA's and AKA's), and the 
major portion of naval forces from 
Halsey's area. 5 

Faced with the possibility of a rival 
offensive, using divisions and ships that 
he had planned to employ, General Mac- 
Arthur hurled back a vigorous reply. 
Arguing against the Central Pacific (he 

1 See JSSC 40/2, 3 Apr 43; JPS 67/4, 28 Apr 43; 
JCS 287, 7 May 43; JCS 287/1. 8 May 43; CCS 220, 
14 May 43. All these papers are entitled "Strategic 
Plan for Defeat of Japan" or something very 
similar. See also Crowl and Love, The Seizure of 
the Gilberts and Marshalls, Ch. I; Smith, The Ap- 
proach to the Philippines, Ch. I; Cline, Wash- 
ington Command Post, Ch. XVII; relevant chapters 
in Morton's forthcoming volumes on strategy, com- 
mand, and logistics in the Pacific, and in Maurice 
MatlofE, Strategic Planning for Coalition Warfare: 
1943-1944, also in preparation for the series 

"Min, JCS mtg, 15 Jun 43; Rad, JCS to Mac- 
Arthur, 15 Jun 43, CM-OUT 6093. 



called the prospective invasion of the 
Marshalls a "diversionary attack"), he set 
forth the virtues of advancing through 
New Guinea to the Philippines. With- 
drawal of the two Marine divisions, he 
maintained, would prevent the ultimate 
assault against Rabaul. He concluded his 
message with the information on target 
dates and forces that the Joint Chiefs 
had requested. 6 Two days later, 22 June, 
Admiral Halsey protested the proposed 
removal of the 2d Marine Division and 
most of his ships. 7 

Although General Mac Arthur may not 
have known it at the time, his argument 
that transfer of the two divisions would 
jeopardize the Rabaul invasion was being 
vitiated. In 1942 there had been general 
agreement that Rabaul should be cap- 
tured, but in June 1943 members of 
Washington planning committees held 
that a considerable economy of force 
would result if Rabaul was neutralized 
rather than captured. 8 The Joint Stra- 
tegic Survey Committee, in expressing 
itself in favor of giving the Central 
Pacific offensive priority over Cart- 
wheel, also argued that the Allied drive 
northward against Rabaul was merely a 
reversal of the Japanese strategy of the 
year before and held "small promise of 
reasonable success in the near future." 9 

On the other hand Admiral William 
D. Leahy, chief of staff to the President 
and senior member of the Joint Chiefs, 

8 Rad, MacArthur to Marshall, 20 Jun 43, CM-IN 

7 Rad, MacArthur to Marshall, 22 Jun 43, CM-IN 
13605. Halsey sent his views to MacArthur who 
relayed them. 

8 Incl B, Jt War Plans Com 58/D, 24 Jun 43, 
title: Memo for Rainbow Team, in OPD 384 Mar- 
shall Islands (10 Jun 43) Sec 1. 

"JSSC 386, 28 Jun 43, title: Strategy of the Pac. 

was always a strong supporter of Mac- 
Arthur's views. 10 He argued strongly 
against any curtailment of Cartwheel. 
Admiral King, however, was far from 
pleased (in June 1943) with the rate 
of "inch by inch" progress in the South 
and Southwest Pacific. He wanted to see 
Rabaul "cleaned up" so the Allies could 
"shoot for Luzon," and seemed to imply 
that if Cartwheel did not move faster 
he would favor a curtailment. 11 

The immediate question on the trans- 
fer of the Marine divisions was com- 
promised. The 1st Marine Division 
would remain in the Southwest Pacific. 
The sd Marine Division, heretofore 
slated for the invasion of Rabaul, was 
transferred from New Zealand to the 
Central Pacific, where it made its bloody, 
valorous assault on Tarawa in Novem- 
ber 1943. Assured by King that the Cen- 
tral Pacific offensive would assist rather 
than curtail Cartwheel, Leahy with- 
drew his objections. 12 

By 2 1 July the arguments against cap- 
turing Rabaul had so impressed General 
Marshall that he radioed MacArthur to 
suggest that Cartwheel be followed by 
the seizure of Kavieng on New Ireland 
and Manus in the Admiralties, with the 

10 See Fleet Admiral William D. Leahy, / Was 
There (New York: Whittlesey House, 1950), passim. 

11 Min, JCS mtg, 29 Jun 43. At this time King 
wanted to go to Luzon by way of the Marianas, 
which he always regarded as the key to the Pacific 
because he believed that an attack there would 
smoke out the Japanese fleet. 

15 Min, JCS mtg, 20 Jul 43; JCS 386/1, 19 Jul 43, 
Strategy in the Pac; JPS 205/3, 10 Jul 43> title: 
Opns Against Marshall Islands; Draft Memo, JPS 
for JCS, 12 Jul 43, sub: Strategy in the Pac, OPD 
381 Security 195; OPD Draft Memo, 14 Jul 43, no 
sub, same file; JPS draft paper, 19 Jul 43, title: 
Strategy in the Pac, and attached papers, with JPS 
219/D in ABC 384 Pac (28 Jun 43); OPD Brief, 
Notes on JWPC 58/2, in OPD 384 Marshall Islands 
(10 Jun 43) Sec 1. 



purpose of isolating Rabaul, and by the 
capture of Wewak. But MacArthur saw 
it otherwise. Marshall's plan, he stated, 
involved too many hazards. Wewak, too 
strong for direct assault, should be 
isolated by seizing a base farther west. 
Rabaul would have to be captured 
rather than just neutralized, he insisted, 
because its strategic location and excel- 
lent harbor made it an ideal naval base 
with which to support an advance west- 
ward along New Guinea's north coast. 13 
Marshall and King were not con- 
vinced. Thus the Combined Chiefs, 
meeting with President Roosevelt and 
Prime Minister Churchill in Quebec 
during August, received and approved 
the Joint Chiefs' recommendation that 
Rabaul be neutralized, not captured. 
They further agreed that after Cart- 
wheel MacArthur and Halsey should 
neutralize New Guinea as far west as 
Wewak, and should capture Manus and 
Kavieng to use as naval bases for sup- 
porting additional advances westward. 
Once these operations were concluded, 
MacArthur was to move west along the 
north coast of New Guinea to the Vogel- 
kop Peninsula. Subsequently MacArthur 
was informed that his cherished ambi- 
tion to return to the Philippines would 
be realized; Marshall radioed him that 
once the Vogelkop was reached, the 
Southwest Pacific's next logical objec- 
tive would be Mindanao. 14 

" Rad 8604, Marshall to MacArthur, 21 Jul 43, 
in Marshall's OUT Log; Rad 16419, MacArthur to 
Marshall, 23 Jul 43, in Marshall's IN Log. 

"See Cline, Washington Command Post, p. 225; 
CCS 319/5, 24 Aug 43, title: Final Rpt to the 
President and Prime Minister; CCS 301/3, 27 Aug 
43, title: Specific Opns in Pac and Far East, 1943- 
44; Rad 8679, Marshal to MacArthur, 2 Oct 43, in 
Marshall's OUT Log. See also Smith, The Approach 
to the Philippines, Ch. I. 

Papers containing the Combined 
Chiefs decisions were delivered to Gen- 
eral MacArthur by Col. William L. 
Ritchie of the Operations Division, War 
Department General Staff, who reached 
GHQ on 17 September. 15 

From then on MacArthur did not raise 
the question of Rabaul with the Joint 
Chiefs; his radiograms dealt instead with 
broader questions relating to the Philip- 
pines and the relative importance of the 
Central and Southwest Pacific offen- 
sives. 16 Although the evidence is not 
conclusive, the general course of events 
and certain opinions MacArthur gave 
during the planning for Bougainville 
seem to indicate that he knew of the 
decision to neutralize rather than cap- 
ture Rabaul, or else had reached the 
same decision independently, some time 
before Colonel Ritchie reached the 
Southwest Pacific. 

The General Plan 

If ever a series of offensives was con- 
ducted according to plan, it was the ex- 
tremely systematic Allied moves in the 
Pacific that started in 1943. At the time 
that Allied forces were fighting in New 
Guinea and New Georgia, the Joint 
Chiefs were considering the wisdom of 
neutralizing Rabaul, and General Mac- 
Arthur and Admiral Halsey were pre- 
paring for the invasion of Bougainville. 

Elkton III had initially provided that 
the southern Bougainville area (Buin 
and Faisi) was to be invaded during the 

"Rad, Ritchie to Handy, 18 Sep 43, CM-1N 

18 MacArthur's subsequent plans called for the 
neutralization of Rabaul, followed by its possible 
capture. See GHQ SWPA, RENO III, so Oct 43, in 
ABC 384 Pac, Sec 3-A. 



fifth month after the beginning of Cart- 
wheel, simultaneously with the con- 
quest of New Georgia, and one month 
before the invasion of Cape Gloucester. 
(See Chart 2.) Admiral Halsey had al- 
tered the plan by managing to start his 
invasion of New Georgia on 550 June. In 
June General MacArthur, in ordering 
the Markham Valley-Huon Peninsula 
attack, directed Admiral Halsey to be 
ready to take southern Bougainville on 
orders from GHQ. 17 At this time Ad- 
miral Halsey, planning in accordance 
with Elkton III, intended to use the 3d 
Marine Division and the 25th Division 
against southern Bougainville, the 2d 
Marine and 3d New Zealand Divisions 
against Rabaul. 18 Before long, however, 
the 25th Division, sent into New Geor- 
gia, was too worn for further combat 
and the 2d Marine Division was ordered 
to invade the Gilberts instead of Ra- 
baul. 19 

Tactical planning for Bougainville be- 
gan in the South Pacific in July when 
Halsey assigned the Commanding Gen- 
eral, I Marine Amphibious Corps, to 
command the ground forces. (Map 15) 

" GHQ SWPA OI 34, 13 Jun 43, in GHQ SWPA 
G— 3 Jnl, 14 Jun 43. This section is based in part on 
Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 245-51; Halsey and Bryan, Admiral 
Halsey's Story, pp. 173—74; Morison, Breaking the 
Bismarcks Barrier, pp. 279-84; Maj. John N. Rentz, 
USMCR, Bougainville and the Northern Solomons 
(Washington, 1948), Ch. I; The Bougainville Cam- 
paign, MS, prepared by Hist Sec G-2, SOPACBA- 
COM, Vol. I, Ch. I, OCMH. None of these, how- 
ever, is entirely satisfactory for they do not employ 
Southwest Pacific Area sources. 

18 Rad, Comdr Third Fit to CINCSWPA, 21 Jun 
43, in GHQ SWPA G- 3 Jnl, 23 Jun 43. 

16 The 25th Division stayed on Guadalcanal after 
the conclusion of the campaign there. It had little 
opportunity for rest and reorganization before 
moving to New Georgia. 

His mission was the seizure of Buin, 
Kahili, and Tonolei Harbor on south- 
ern Bougainville and of the nearby is- 
lands in Bougainville Strait— the Short- 
lands, Faisi, and Ballale, where there 
were then an estimated twenty thousand 
Japanese soldiers and sailors. 

Near the end of July Admiral Halsey 
suggested a change in plan to General 
MacArthur. It was based on two assump- 
tions: first, that the objectives of the 
operation were denying the use of air- 
fields and anchorage to the Japanese 
and securing airfields and anchorages 
for the Allies, as a step toward the cap- 
ture of Rabaul; and second, that because 
terrain, strategic position, and Japanese 
dispositions indicated that southern Bou- 
gainville was extremely important to 
the Japanese, the operation would be 
a major one. With the difficulties of the 
then bogged-down New Georgia inva- 
sion and the success of the artillery on 
the offshore islands against Munda both 
obviously in mind, he suggested that 
he could save men, materiel, and time 
by avoiding the Bougainville mainland 
completely. He proposed to seize the 
Shortlands and Ballale, to emplace artil- 
lery on the former with the mission of 
interdicting Kahili, to build one or more 
airfields in the Shortlands, and to use 
the anchorages there that the Japanese 
8th Fleet then employed regularly. Mac- 
Arthur heartily approved the scheme. 20 

By early September, however, Admiral 
Halsey had decided on a further change 

^Ltr, COMSOPAC to CINCSWPA, 26 Jul 43, 
sub: Bougainville Opn, in GHQ SWPA G-3 Jnl, 
31 Jul 43; Rad, CINCSWPA to Comdr Third Fit, 
4 Aug 43, in GHQ SWPA G- 3 Jnl, 4 Aug 43; 
COMSOPAC War Diary, a6 Jul and 4 Aug 43 

» \tt Green Is 


07 J 6^.^ 

I NOV jfi<^£SIZ 
S Cope" 
3 l+) Torokina 


0432 * Mor empxfss 


27 October-1 November 1943 

p. Allied landing date indicated 



1000 S0O0 5000 AND ABOVE 
10 20 30 10 MILES 

10 20 JO 40KIL0WETERS 







Z ygyiPrcht 



9 ! 

Stirling I 



LAVE L LA \ ( /'"V^ ^ 

Printed by Army Map Service 



in plan, Several factors influenced his 
decision. The impressive and inexpen- 
sive success on Vella Lavella had demon- 
strated once more the validity of the old 
principle of striking soft spots, when 
possible, in preference to headlong as- 
sault against fixed positions. Further, 
reconnaissance had indicated that air- 
drome sites on the Shortlands were not 
very good. Landing in the Shortlands, 
which the Japanese were believed to 
be reinforcing, would entail heavy losses; 
poor beaches would impede the landing 
of heavy construction equipment and 
artillery for the neutralization of Kahili. 
It was also estimated that assaulting the 
Shortlands-Ballale-Faisi area would re- 
quire two divisions, while two more 
would be needed to operate on southern 
Bougainville proper. As the South Pa- 
cific had but four divisions— the 37th 
and Americal Divisions of the U.S. 
Army, the 3d Marine Division, and the 
3d New Zealand Division— that were 
considered fit to fight, no more advances 
would be possible for months. 21 

Looking for a method of neutralizing 
the southern Bougainville-Shortlands 
area without capturing it, a method that 
would retain enough troops for a major 
forward move later, Halsey acted on the 
advice of his principal subordinate com- 
manders. He decided in favor of in- 
creased air effort from the New Georgia 
fields against southern Bougainville and 
Buka. Starting about 1 November, he 
proposed to capture the Treasury Islands 
and Choiseul Bay as airfield, radar, and 
PT base sites from which to "contain 

" The ad Marine Division was due to leave; the 
25th and 43d Divisions were due for rest and re- 

and strangle" southern Bougainville and 
the Shortlands. He proposed that after 
the mainland of Bougainville had been 
reconnoitered he and MacArthur could 
decide whether to advance from Choi- 
seul to Kieta on the east coast or from 
the Treasuries to Empress Augusta Bay 
on the west if post-CARTWHEEL plans 
required the establishment of positions 
on the mainland of Bougainville. 22 

This plan was consistent with Elkton 
III, and varied only slightly from the 
July schemes approved by MacArthur. 
But by now, MacArthur, perhaps aware 
of the decision to neutralize rather than 
capture Rabaul, and obviously anxious 
to hurry up Cartwheel and get started 
on the drive toward the Philippines, 
had changed his mind about the scope 
and nature of the operation. Thus when 
Halsey's chief of staff, Rear Adm. Rob- 
ert B. Carney, and his new war plans 
officer, Col. William E. Riley, USMC, 
presented the Treasuries-Choiseul plan 
to MacArthur at GHQ on 10 September, 
MacArthur was against it. With the suc- 
cessful airborne move to Nabzab in 
mind, he expressed his agreement with 
the principle of the bypass, but main- 
tained that Halsey's plan would make it 
impossible for South Pacific aircraft to 
hit at Rabaul effectively before 1 March 

"Ltr, Halsey to CINCSWPA, 9 Sep 43, sub: 
Elkton III— S Bougainville Objectives, in GHQ 
SWPA G-3 Jnl, 10 Sep 43; Memo, Adm Fitch, Gen 
Harmon, Maj Gen Charles D. Barrett [CG I Mar 
Amphib Corps], and Adm Wilkinson for COM- 
SOPAC, 7 Sep 43, no sub, ABC 384 (1-17-43) Sec 
2; Halsey, Narrative Account of the South Pacific 
Campaign, p. 8, OCMH; Harmon, The Army in the 
South Pacific, p. 9, OCMH. Some advocated by- 
passing Bougainville completely in favor of a jump 
to Emirau in the Saint Matthias group northwest 
of Kavieng. 



1944. He wanted Halsey's aircraft estab- 
lished within fighter range of Rabaul in 
time to assist with the neutralization of 
Rabaul that would cover the Southwest 
Pacific's invasion of Cape Gloucester. 
This would be necessary, MacArthur 
held, because Southwest Pacific air forces 
could not attack all the objectives (in- 
cluding Madang and Wewak) that would 
have to be neutralized in order to pro- 
tect the invasions of Cape Gloucester 
and of Saidor, on the north coast of the 
Huon Peninsula, Southwest Pacific 
headquarters hoped to start Operation 
III (chiefly Cape Gloucester) shortly 
after 1 December; Cape Gloucester it- 
self would probably be invaded between 
25 December 1943 and 1 January 1944. 
Therefore it would be necessary for 
South Pacific forces to establish them- 
selves on the mainland of Bougainville 
about 1 November. So important was 
the operation that MacArthur tacitly 
approved commitment of the major part 
of South Pacific ground forces. 

Specifically, he proposed the following 
outline plan: 

1. 15 October- 1 November, South- 
west Pacific air forces would make heavy 
attacks against Japanese aircraft, air in- 
stallations, and shipping at Rabaul; 

2. 20-25 October, South Pacific forces 
would occupy the Treasuries and posi- 
tions on northern Choiseul in order to 
establish radar positions and PT boat 

3. 1 November, South Pacific forces 
would occupy Empress Augusta Bay on 
the west coast of Bougainville in order 
to establish airfields within fighter range 
of Rabaul; 

4. 1-6 November, the Southwest Pa- 

cific would continue air attacks on Ra- 
baul and would assist in the neutraliza- 
tion of Buka; 

5. 25 December 1943-1 January 1944, 
Southwest Pacific forces would seize 
Cape Gloucester and Saidor in order to 
gain control of Vitiaz and Dampier 
Straits and to secure airdromes for the 
neutralization of Kavieng. During this 
period South Pacific forces would neu- 
tralize Rabaul. 23 

General MacArthur stressed the im- 
portance of a landing on the mainland 
at another meeting on 17 September 
attended by General Harmon and Col- 
onel Riley. Asked if he preferred a land- 
ing on the east or the west coast of Bou- 
gainville, he put the decision entirely 
in Admiral Halsey's hands. 

And so on 22 September, Halsey is- 
sued warning orders which canceled all 
his earlier plans and assigned the units 
to constitute the invasion force. Ad- 
miral Wilkinson would lead it. The 
landing forces, under Wilkinson, were 
still to be under the commanding gen- 
eral of the I Marine Amphibious Corps. 
Halsey instructed Wilkinson and his 
units to be ready to carry out one of 
two plans: either they were to seize and 
hold the Treasury Islands and the air- 
field sites in the Empress Augusta Bay 
region on the west coast of Bougainville; 
or they were to seize the Treasuries and 
Choiseul Bay, build airfields, PT boat 
bases, and landing craft staging points, 
and in late December seize the Japanese 

23 Ltr, MacArthur to Halsey, 1 1 Sep 43, no sub, 
and Notes for Memo on Conf Between Reps of 
SWPA and SOPAC, GHQ SWPA, 10 Sep 43. Both 
in GHQ SWPA G-3 Jnl, 1 1 Sep 43. 



airfield at Tenekau on the east coast of 
Bougainville. 24 

Submarines took patrols to the east 
coast and to Empress Augusta Bay to 
gather data, and South Pacific intelli- 
gence officers interviewed missionaries, 
traders, planters, coastwatchers, and 
fliers who had been shot down over 
Bougainville. The east coast patrol, car- 
ried by the submarine Gato, delivered 
an unfavorable report. The west coast 
patrol, composed of marines, debarked 
from the submarine Guardfish about ten 
miles northwest of Cape Torokina in 
Empress Augusta Bay. The marines were 
unable to examine Cape Torokina be- 
cause it was occupied by the Japanese, 
but they took samples of soil similar to 
that at Torokina. When tested, it showed 
that Cape Torokina was suitable for 

Between the sea and the mountains 
at Cape Torokina, which lay within 
fighter range of Munda, was a coastal 
plain of about seven square miles. It 
was lightly defended; Halsey estimated 
that there were about one thousand Jap- 
anese in the area. So forbidding were 
the surrounding mountains that the area 
was almost isolated from the strong Jap- 
anese garrisons in southern Bougainville. 
Halsey and his planners estimated that 
if Allied forces seized Torokina the 

M Ltr, COMSOPAC to CG 1 Mar Amphib Corps, 
CTF 31, and CTF 33, 22 Sep 43, sub: Warning 
Order, in GHQ SWPA G-3 Jnl, 24 Sep 43. During 
this period Admiral Halsey received communica- 
tions from Admiral King's office which seemed to 
require him to seize southern Bougainville and then 
Kieta and Buka. This confused the issue until Ad- 
miral Nimitz assured Halsey that the messages from 
King were estimates and not directives, and that 
Halsey was to operate under the provisions of the 
28 March 1943 directive. 

Japanese would require three or four 
months to bring enough heavy equip- 
ment over the mountains to launch an 
effective counterattack. But there were 
disadvantages. The heavy surf in Em- 
press Augusta Bay, which had no pro- 
tected anchorages, would make landing 
operations difficult. No more than 65 
miles separated the cape from all the 
Japanese air bases on Bougainville, and 
Rabaul was only 215 miles to the north- 

Admiral Halsey calculated the chances 
and decided on Torokina. In his words: 
"The conception was bold and the prob- 
ability of provoking a violent air-land- 
surface action was accepted and wel- 
comed On the premise that the by-prod- 
ucts of enemy destruction would, in 
themselves, greatly further the over-all 
Pacific plan. Enthusiasm for the plan 
was far from unanimous, even in the 
South Pacific, but, the decision having 
been made, all hands were told to 'Get 
going.' " 25 

Halsey informed MacArthur of his 
decision on 1 October. Expressing his 
complete agreement, MacArthur prom- 
ised maximum air support from the 
Southwest Pacific. The invasion would 
be launched on 1 November. 26 

Air Operations in October 

The Fifth Air Force 

By October the Fifth Air Force in 
the Southwest Pacific Area was well situ- 

M Halsey, Narrative Account of the South Pacific 
Campaign, p. 8, OCMH. 

" Rad, Halsey to MacArthur, 1 Oct 43, and Rad, 
MacArthur to Halsey, 1 Oct 43, in GHQ SWPA G-3 
Jnl, 1 Oct 43. 



ated to carry the fight against Rabaul. 21, 
Nearly all its warplanes had been dis- 
placed to forward bases. Port Moresby, 
an outpost in 1942, was now a rear base. 
Dobodura was the main staging base 
for heavy bombers, and Nadzab was 
being readied as the main base for fu- 
ture operations. P-38's from New 
Guinea could stage through Kiriwina 
and escort the bombers all the way to 

Rabaul was ripe for air attack. Trans- 
ports, cargo ships, and smaller craft, to- 
gether with some warships, crowded 
Simpson Harbor. Supply depots were 
fully stocked. Four all-weather airfields 
— Lakunai, Vunakanau, Rapopo, and 
Tobera— were in operation in and near 
Rabaul. 28 

Southwest Pacific aircraft had been 
harrying Rabaul with small raids since 
January 1942, but now the Allies were 
ready to attack this bastion on a large 
scale. General Kenney was ready for 
the first big attack on 12 October. All 
together, 349 planes took part: 87 heavy 
bombers, 114 B-25's, 12 Beaufighters, 
and 125 P-38's, plus some weather and 
photo reconnaissance planes— or, as he 
put it, "Everything that I owned that 

r This section is based on Craven and Cate, The 
Pacific: Guadalcanal to Saipan, pp. 851-55, 317-26; 
Kenney, General Kenney Reports, 313—20; Morison, 
Breaking the Bistnarcks Barrier, pp. 275—88, 271-92; 
Thomas C. Wilds, "The Admiral Who Lost His 
Fleet," United States Naval Institute Proceedings, 
Vol. 77, No. 11 (November 1951); CTF 33 War 
Diary, Oct 43 entries; Southeast Area Naval Opera- 
tions, Vol. Ill, Japanese Monogr No. 50 (OCMH), 
pp. 5-11; Outline of Southeast Area Naval Air 
Operations, Pt. IV, Japanese Monogr No. 108 
(OCMH), pp. 42-44. 

2 * Lakunai had a sand and volcanic ash surface; 
the other three were concrete. Keravat field on the 
west coast of Gazelle Peninsula had never been 

was in commission, and could fly that 
far." 28 B-25's and Beaufighters made 
sweeps over Vunakanau, Rapopo, and 
Tobera while the heavy bombers struck 
at shipping. The Allies lost four planes 
and estimated a great deal of damage to 
Japanese aircraft and ships. Their esti- 
mates were somewhat exaggerated, espe- 
cially those on shipping damage, but 
some Japanese planes were destroyed. 
The Japanese, taken by surprise and 
unable to send up fighters to intercept, 
later reported that this and later raids 
in October were "a great obstacle to the 
execution of operations." 30 

Bad weather over New Guinea halted 
Kenney's operations against Rabaul for 
the next few days. The Japanese used 
the respite to send out attacks against 
Oro Bay on 15 and 17 October, and 
Finschhafen on 17 and 19 October. The 
Allied planes did not sit idle while Ra- 
baul was inaccessible, but struck at We- 
wak on the 16th and again the next day. 

Kenney planned and sent out another 
big raid against Rabaul on 18 October, 
but when the air armada was over the 
Solomon Sea the weather closed in. 
Fifty-four B-25's went on to Rabaul 
anyway. Kenney followed this attack 
with three successive daylight raids on 
23, 24, and 25 October before the 
weather again imposed a delay, this time 
until the 29th, when B-24's and P-38's 
bombed Vunakanau. 

The weather intervened again, so that 
it was not until 2 November, the day 
after South Pacific forces landed at Em- 

** Kenney, General Kenney Reports, p. 313. 

30 Outline of Southeast Area Naval Air Opera- 
tions, Pt. IV, Japanese Monogr No. 108 (OCMH), 
p. 44. 

Bombing Rabaul. Top left and right: Japanese corvette off the coast of New 
Britain near Rabaul suffers a direct hit from a B-25. Bottom: Vunakanau airfield 
is attacked by low-flying bombers dropping parachute bombs. 



press Augusta Bay, that Southwest Pacific 
aircraft again struck at Rabaul. On that 
day seventy-five B-25's escorted by P-38's 
attacked and ran into the fiercest oppo- 
sition the Fifth Air Force encountered 
during World War II. A large number 
of carrier planes and pilots from the 
Combined Fleet at Truk had just been 
transferred to Rabaul, and they put up 
a stiff fight. 

Although it is clear that these raids 
failed to wreak as much havoc at Rabaul 
as Kenney's fliers claimed, it is also clear 
that they caused a good deal of damage 
to aircraft and prevented the Japanese 
planes at Rabaul from undertaking any 
purely offensive missions. In short, the 
Southwest Pacific's air support for the 
Bougainville invasion, though not as 
devastating as was thought at the time, 
was effective. 

Certainly American pilots, like the 
Japanese, and like soldiers and sailors on 
the ground and in ships, tended to exag- 
gerate the damage they inflicted. But 
there were two important differences 
between American and Japanese claims. 
First, Japanese claims were wildly exag- 
gerated whereas American claims were 
merely exaggerated. Second, Japanese 
commanders apparently took the claims 
seriously, so that nonexistent victories 
often served as the bases for decision. On 
the other hand American commanders, 
taking human frailty into account, eval- 
uated and usually scaled down claims so 
that decisions were normally based on 
more realistic estimates of damage. 

Air Command, Solomons 

General Twining's composite force, 
Air Command, Solomons, had been strik- 
ing hard at the northern Solomons bases 

during the same period and for the 
same purpose— to knock out the Bou- 
gainville bases so that Wilkinson's con- 
voys could sail past in safety. Twining's 
available air strength had been displaced 
forward to bases within range of south 
Bougainville targets. At the start of op- 
erations in October, Twining had 614 
Army, Navy, Marine Corps, and Royal 
New Zealand Air Force planes. Of these, 
264 fighters and 223 medium, dive, tor- 
pedo, and patrol bombers were at New 
Georgia and the Russells, and 127 heavy 
bombers and patrol planes were at Guad- 

Ever since 1942 South Pacific planes 
had been battering at the Japanese bases 
at Kahili, the Shortlands, Ballale, Kieta, 
and Buka, and now the process was in- 
tensified in an effort to put them out of 
commission. 31 Starting on 18 October, 
Twining— whose high professional quali- 
fications were matched by a physical ap- 
pearance so striking that he looked like 
Hollywood's idea of a diplomat— drove 
his interservice, international force hard 
in a continuous series of high-level, low- 
level, dive, glide, and torpedo bombing 
attacks and fighter sweeps, all made with 
escorting fighters from the four air serv- 
ices in the command. The primary mis- 
sion was accomplished. The hard-hit 
enemy showed skill and determination 
in keeping his airfields in repair, but 
these qualities were not enough. By 1 
November all his Bougainville airfields 
had been knocked out of commission, 
and the continuous attacks kept them 
that way. 

" Kenney offered to include Buka in his attacks, 
but Halsey asked him to concentrate on Rabaul and 
leave Buka to Twining. 



B-25's Leaving Bougainville after an attack on airfields and supply areas. 

The Japanese 

Of Admiral Kusaka's nth Air Fleet, a 
substantial portion was based at Rabaul 
in early October, the remainder in south- 
ern Bougainville. When Air Command, 
Solomons, intensified its operations, Ku- 
saka withdrew his planes to Rabaul, and 
to avoid being completely destroyed by 
Kenney's heavy raids he frequently 
pulled his planes back to Kavieng in New 
Ireland. Despite these attacks Kusaka was 
usually able to maintain about two hun- 
dred planes in operating condition at 
Rabaul throughout October. 

Now Admiral Koga, like the late Ya- 
mamato, decided to use his carrier planes 
jointly with the land-based planes of 
Kusaka's air fleet in an effort to improve 

the situation in the Southeast Area. As 
a result of the September decision to 
withdraw the main defensive perimeter, 
Koga developed a plan to cut the Allied 
lines of communication in the Southeast 
Area and so delay the Allies and buy 
time for the Japanese to build up the 
defenses along the main perimeter. This 
plan, called Operation RO, was to be 
executed by the operational carrier air 
groups of the Combined Fleet, trans- 
ferred from Truk to Rabaul, and by the 
nth Air Fleet. Vice Adm. Tokusaburo 
Ozawa, commander of the 3d Fleet, and 
Kusaka would conduct the operation 
jointly from Rabaul, Koga decided on 
this course of action fully aware that his 
surface strength would be immobilized 



while his carrier planes were at Rabaul. 

He had planned to transfer the planes 
in mid-October, but delayed the move 
because he received a false report that 
the U.S. Pacific Fleet was out against 
the Marshalls. On 20 October, now aware 
that Nimitz' forces were not moving 
against the Marshalls, Koga ordered the 
carrier planes dispatched. By the begin- 
ning of November, 173 carrier planes— 
82 fighters, 45 dive bombers, 40 torpedo 
bombers, and 6 patrol planes— had 
reached Rabaul to team with Kusaka's 
200. It was Ozawa's carrier pilots who 
gave Kenney's men such a hard fight on 
2 November. Koga had first planned to 
deliver his main stroke against New 
Guinea but the increased tempo of Al- 
lied activity in the Solomons made him 
decide to strike in the Solomons. 

Koga's decision to execute Operation 
RO was to have far-reaching results, re- 
sults that were the precise opposite of 
what he expected. The transfer of the 
carrier planes coincided with the South 
Pacific's invasion of Bougainville. 

Forces and Tactical Plans 

The Allies 

Bougainville, the largest of the Solo- 
mon Islands, is 125 miles long on its 
northwest-to-southeast axis, and 30 to 48 
miles wide. 32 Its mountainous spine com- 

32 This section is based on Morison, Breaking the 
Bismarcks Barrier, pp. 279-89; Halsey and Bryan, 
Admiral Halsey's Story, pp. 173-76; SOPACBACOM, 
The Bougainville Campaign, Vol. I, Ch. I, OCMH; 
Japanese Operations in the Southwest Pacific Area 
(Vol. II of MacArthur hist), Ch. IX, p. 8i6, OCMH; 
Southeast Area Naval Operations, III, Japanese 
Monogr No. 50 (OCMH), 7-8; 8th Area Army 
Operations, Japanese Monogr No. 110 (OCMH) 
pp. 85-103; 17th Army Operations, II, Japanese 
Monogr No. 40 (OCMH), 95-99; COMSOPAC Opn 
Plan 16-43, 1! * Oct 43, in GHQ SWPA G-3 Jnl, 17 
Oct 43. 

prises two ranges, the Emperor and the 
Crown Prince. Two active volcanoes, 
10,000-foot Mount Balbi and 8,000-foot 
Mount Bagana, send continual clouds 
of steam and smoke into the skies. Mount 
Bagana, a stark and symmetrical cone, 
overlooks Empress Augusta Bay and is 
the most outstanding feature of the re- 
gion's dramatic beauty. 

The mountain range ends toward the 
southern part of the island, and there, 
on the coastal plain near Buin, the Jap- 
anese had built the airfields of Kahili 
and Kara. On the western coast the 
mountains slope down through rugged 
foothills and flatten out into a narrow 
and swampy coastal plain that is cut 
by many small rivers. These silt-laden 
streams constantly build bars across their 
own mouths and thus frequently change 
their courses. 

Good harbors in varying stages of 
development were to be found at Buka, 
Numa Numa, Tenekau, Tonolei, and 
in the islands off the south coast. Em- 
press Augusta Bay, exposed as it was to 
the open sea, was a poor anchorage. The 
Japanese had airfields at Buka and Bonis 
on either side of Buka Passage, at Tene- 
kau, Kieta, Kara, and Kahili on the 
mainland, and at Ballale near the Short- 
lands, and had seaplane anchorages and 
naval bases in the Shortlands. As on all 
the other islands, there were no real 
motor roads, only native trails near the 
coasts plus a few that led through the 

The native population consisted of 
over forty thousand nominally Christian 
Melanesians, who were slightly darker 
in color than their fellows in the south- 
ern Solomons. Before the war about a 
hundred white missionaries, planters, 



traders, and government officials had 
lived on the island. Some of the natives, 
it was known, were pro-Japanese and 
had aided the enemy in rooting out the 
coastwatchers earlier in the year. 

Allied intelligence agencies estimated 
enemy strength at about 37,500 soldiers 
and 20,000 sailors, and correctly reported 
that the Army troops belonged to the 
lyth Army, commanded by General 
Hyakutake, who had been responsible 
for the direction of the Guadalcanal 
Campaign. 33 Over 25,000 of Hyakutake's 
men were thought to be in the Buin- 
Shortlands area, with an additional 
5,000 on the east coast of Bougainville, 
5,000 more at Buka and Bonis, and light 
forces at Empress Augusta Bay. Air re- 
connaissance enabled the Allies to keep 
a fairly accurate count of Japanese war- 
ships and planes in the New Guinea- 
Bismarcks-Solomons area. 

Admiral Halsey, in preparing his at- 
tack, was not embarrassed by too many 
ships. Admiral Nimitz was getting ready 
to launch his great Central Pacific ad- 
vance in November and had removed 
many of Halsey's ships, leaving him but 
eight transports and four cargo ships, or 
enough shipping to carry one reinforced 
division in the assault. Because South 
Pacific commanders expected the Jap- 
anese to oppose the invasion with vigor- 
ous air attacks, they decided not to use 
the slow LST's for the assault. The 
South Pacific had one carrier force, Task 
Force 38 under Rear Adm. Frederick C. 
Sherman, consisting of the 910-foot air- 

craft carrier Saratoga, the light carrier 
Princeton, two antiaircraft cruisers, and 
ten destroyers. Nimitz, in response to 
Halsey's requests for additional cruiser- 
destroyer and carrier task forces, as- 
sured Halsey that Central Pacific forces 
would be within reach to assist if neces- 
sary, and agreed to send Halsey another 
carrier task force on or about 7 Novem- 
ber. 34 

Halsey issued the basic orders for the 
operation on 12 October. He organized 
five task forces similar to those that had 
made up the New Georgia attack forces. 
They were: Task Force 31 (the attack 
force), under Admiral Wilkinson; Task 
Force 33 (South Pacific land-based air- 
craft), under Admiral Fitch; Sherman's 
Task Force 38; the cruisers and destroy- 
ers of Admiral Merrill's Task Force 39; 
and Captain Fife's submarines in Task 
Force 72. 

The submarines were to carry out of- 
fensive reconnaissance in the waters of 
the Bismarck Archipelago, and would be 
supplemented in their work by Central 
Pacific submarines operating out of Pearl 
Harbor. Merrill's ships would support 
the invasion by operating against enemy 
surface ships and by bombarding Buka 
and the Shortlands. Halsey also planned 
to employ Sherman's Task Force 38 in 
a raid against Buka and Bonis, which 
lay beyond effective fighter range of the 
New Georgia airfields. Task Force 33 
was ordered to carry out its usual mis- 
sions of reconnaissance, destruction of 
enemy ships and aircraft, and air cover 
and support of the invasion force. Air 

33 "Harukichi," listed in Miller, Guadalcanal: The 

First Offensive as Hyakutake's given name, is a mis- "Ltr, Adm Duncan to Gen Smith, Chief of Mil 
translation. Hist, 16 Nov 53, no sub, OCMH. 



Command, Solomons, which was part 
of Task Force 33, was making its inten- 
sive effort during October against the 
Japanese airfields in southern Bougain- 
ville and the outlying islands, so that 
these areas could safely be bypassed. Ar- 
rangements for local air support were 
the same as for New Georgia. The local 
air commander with the invasion force 
was designated, as a subordinate of 
Twining's, Commander, Aircraft, North- 
ern Solomons, and would take command 
of all support aircraft as they took off 
from their bases. 

Admiral Wilkinson's invasion force, 
Task Force 31, consisted of eight trans- 
ports, four cargo ships, two destroyer 
squadrons, mine craft, almost all the 
South Pacific's PT squadrons, and a 
large force of ground troops under the 
Commanding General, I Marine Am- 
phibious Corps (I MAC). 

The ground commander was General 
Vandegrift, USMC, an apple-cheeked, 
deceptively soft-spoken Virginia gentle- 
man, who had won distinction by his 
conduct of operations on Guadalcanal 
from 7 August 1942 until December of 
that year. Vandegrift was at this time 
slated to become commandant of the 
Marine Corps in Washington, but was 
given the Bougainville command tem- 
porarily because Maj. Gen. Charles D. 
Barrett, who had replaced Vogel in com- 
mand of the I Marine Amphibious 
Corps, had met accidental death in Nou- 
mea. Halsey's choice for the corps com- 
mand fell upon Maj. Gen. Roy S. Geiger, 
USMC, another hero of Guadalcanal, 
who was then in Washington as Director 
of Marine Corps Aviation. Vandegrift 
was to exercise the command until 
Geiger could arrive. 

Ground forces assigned to the attack 
included the following: I Marine Am- 
phibious Corps headquarters and corps 
troops; 3d Marine Division; 37th Divi- 
sion; 8th Brigade Group, 3d New Zea- 
land Division; 3d Marine Defense Bat- 
talion; 198th Coast Artillery Regiment 
(Antiaircraft); 2d Provisional Marine 
Raider Battalion; 1st Marine Parachute 
Battalion; naval construction and com- 
munications units, and a boat pool. 

In area reserve, to be committed on 
orders from Admiral Halsey, were the 
Americal Division in the Fijis; the 2d 
Battalion, 54th Coast Artillery (Harbor 
Defense) Regiment at Espiritu Santo; 
and the 251st Coast Artillery (Antiair- 
craft) Regiment in the Fijis. 

Naming D Day as 1 November, the 
date for the invasion of Empress Augusta 
Bay, Halsey ordered Task Force 31 to 
seize and hold the Treasury Islands on 
D minus 5 (27 October) and establish 
radar positions and a small naval base. 
Wilkinson's main attack would be the 
seizure of Empress Augusta Bay on 1 
November, which would be followed by 
the speedy construction of two airfields 
on sites to be determined by ground re- 
connaissance after the troops had landed. 
Task Force 31 was initially ordered to 
be ready to establish a PT base on north- 
ern Choiseul. This part of the plan was 
changed on the recommendation of 
Vandegrift, who argued that the Treas- 
ury landings might reveal to the Jap- 
anese the intention to invade Empress 
Augusta Bay. Halsey, Wilkinson, and 
Vandegrift decided instead to use the 
2d Marine Parachute Battalion in a 
twelve-day raid on Choiseul which they 
hoped would mislead the enemy into 



believing that the real objective lay on 
Bougainville's east coast. 35 

Halsey made Wilkinson, whose head- 
quarters was then at Guadalcanal, re- 
sponsible for co-ordination of all am- 
phibious plans. Wilkinson was to com- 
mand all elements of Task Force 31 
until, at a time agreed upon by him and 
the ground commander, direction of all 
air, ground, and naval forces at Empress 
Augusta Bay would be transferred to the 

Wilkinson divided Task Force 31 into 
a northern force, which he commanded 
himself, for the main attack and a south- 
ern force, led by Admiral Fort, for the 
Choiseul raid and the seizure of the 
Treasuries. The assault echelon of the 
northern force, scheduled to land at Em- 
press Augusta Bay on D Day, included 
destroyers, the transports and cargo ships, 
and Maj. Gen. Allen H. Turnage's 3d 
Marine Division, less one regimental 
combat team and plus supporting units. 
The Treasuries echelon of the southern 
force was made up of 8 APD's, 2 LST's, 
8 LCI's, 4 LCT's, 2 APC's, the 8th Bri- 
gade Group of the 3d New Zealand Di- 
vision, the 198th Coast Artillery, A Com- 
pany of the 87th Naval Construction Bat- 
talion, and communications and naval 
base detachments. The parachute bat- 
talion would be transported by four 
APD's escorted by destroyers. The 37th 
Division, in corps reserve, would be 
picked up at Guadalcanal by the north- 
ern force transports and would start ar- 

51 General Geiger described the plan of maneuver 
as "a series of short right jabs designed to throw 
the enemy off balance and conceal the real power 
of the left hook to his midriff at Empress Augusta 
Bay." He must have boxed left-handed. 

Lt, Gen. Alexander A. Vandegrift 

riving at Bougainville soon after D Day 
to help hold the beachhead. 

Guadalcanal and the Russells were to 
serve as the main staging and supply 
bases. However, the shortage of shipping 
led the I Marine Amphibious Corps to 
shorten the lines by establishing a supply 
base at Vella Lavella. Plans called for 



the Vella depot to be stocked with a 
thirty-day supply of rations and petro- 
leum products, but so strained was South 
Pacific shipping that only a ten-day sup- 
ply had been stocked at Vella Lavella 
by 1 November. 

During the last half of October the 
ground units completed their training 
and conducted final rehearsals. The 3d 
Marine Division, part of which had 
served in Samoa in 1942 before joining 
the main body in New Zealand, had re- 
cently transferred from New Zealand to 
Guadalcanal. It completed its amphibi- 
ous and jungle training there and re- 
hearsed for Empress Augusta Bay in the 
New Hebrides from 16 to 20 October. 
The 37th Division, returned from New 
Georgia to Guadalcanal in September, 
likewise conducted amphibious and jun- 
gle training at Guadalcanal. The 3d Ma- 
rine Defense Battalion, which after serv- 
ing in the Guadalcanal Campaign had 
been sent to New Zealand and from 
there back to Guadalcanal, rehearsed 
there. The 8th Brigade practiced land- 
ings at Efate en route to Guadalcanal 
from New Caledonia, and from 14 to 17 
October rehearsed at Florida. 

The Japanese 

The Japanese fully expected Halsey to 
attack Bougainville and were busy pre- 
paring to meet the invasion. Imperial 
Headquarters' orders in September had 
stressed the importance of Bougainville 
as an outpost for Rabaul, and General 
Imamura had instructed General Hya- 
kutake to make ready. This the ijth 
Army commander did, acting in con- 
junction with the commander of the 
8th Fleet. The Japanese planned to use 

air and surface strength to smash any 
Allied attempt at invasion before the 
assault troops could get off their trans- 
ports. But if troops did succeed in get- 
ting ashore, the Japanese hoped to at- 
tack and destroy their beachheads. 

Hyakutake's army consisted mainly of 
the 6th Division, Lt. Gen. Masatane 
Kanda commanding. (This division had 
acquired an unsavory reputation for in- 
discipline by its sack of Nanking, China, 
in 1937). Also assigned were the 4 th 
South Seas Garrison Unit (three infan- 
try battalions and one field artillery bat- 
tery), and field artillery, antiaircraft ar- 
tillery, and service units. Imamura was 
sending four rifle battalions and one 
artillery battalion of the ijth Division 
from New Britain to northern Bougain- 
ville; these were due in November. 36 

Hyakutake, whose headquarters was 
on tiny Erventa Island near Tonolei 
Harbor, had disposed most of his strength 
to cover the Shortlands, Buin, and To- 
nolei Harbor, the rest to protect Kieta 
and Buka. Some 26,800 men— 20,000 of 
the ijth Army and 6,800 of 8th Fleet 
headquarters and naval base forces— and 
an impressive number of guns ranging 
from machine guns to 140-mm. naval 
rifles were stationed in southern Bou- 
gainville and the islands. Over 4,000 
men were at Kieta, and the arrival of 
the ijth Division units would bring the 
Buka Passage garrison to 6,ooo. 37 

M The units attached to the Southeastern Detach- 
ment had been returned to their parent organiza- 
tions. The Detachment was inactivated in December. 

31 These figures are taken from Southeast Area 
Naval Operations, III, Japanese Monogr No. 50 
(OCMH), 7—8. Army accounts do not give strength 



The unpromising nature of the ter- 
rain on the west coast of Bougainville 
had convinced Hyakutake that the Al- 
lies would not attempt to land there. 
Consequently only a small detachment 
was stationed at Empress Augusta Bay. 
Hyakutake was aware that he would be 
outnumbered and outgunned in any 
battle, but like most of his fellow Jap- 
anese generals he placed great faith in 
the superior morale he believed his 
troops possessed. 

"The battle plan is to resist the en- 
emy's material strength with persever- 
ance, while at the same time displaying 
our spiritual strength and conducting 
raids and furious attacks against the 
enemy flanks and rear. On this basis we 
will secure the key to victory within 
the dead spaces produced in enemy 
strength, and, losing no opportunities, 
we will exploit successes and annihilate 
the enemy." 38 Pride goeth before de- 
struction, and an haughty spirit before 
a fall. 

Preliminary Landings 
The Treasuries 

The assault echelon of Admiral Fort's 
southern force consisted of five transport 
groups: the advance transport group 
with 8 APD's and 3 escorting destroyers; 
the second with 8 LCI(L)'s, 2 LCI(G)'s, 39 
and 6 destroyers; the third with 2 LST's, 
2 destroyers, and 2 minesweepers; the 
fourth with 1 APC, 3 LCT's, and 2 PT 
boats; the fifth with 1 APC, 6 LCM's, 

" 8 17th Army Operations, H, Japanese Monogr No. 
40 (OCMH), 95. 

39 The LCI (G) was a gunboat designed to give 
close fire support in landings. Two ao-mm., three 
40 mm., and five .50-caliber machine guns were in- 
stalled on an LCI (L). 

and a rescue boat. 40 These ships loaded 
troops and supplies at Guadalcanal, Ren- 
dova, and Vella Lavella and departed 
for the Treasuries on 26 October. Their 
departures were timed for the five groups 
to arrive in Blanche Harbor, which is 
between Mono and Stirling Islands, be- 
tween 0520 and 0830, 27 October. 41 All 
possible measures were taken to avoid 
detection, because the small forces had 
to get established in the Treasuries be- 
fore the Japanese were able to send in 
reinforcements from their ample re- 
serves in the nearby Shortlands. But 
detection was almost inevitable in an 
operation so close to enemy bases, and 
at 0420, 27 October, a reconnaissance 
seaplane sighted the ships near the 
Treasuries and reported their presence. 
Admiral Merrill's task force, covering 
the operation some distance westward, 
was also discovered. 

Heavy rain fell as the leading APD's 
arrived off the western entrance to 
Blanche Harbor. Low-hanging clouds 
obscured the jungled hills of Mono Is- 
land. As Blanche Harbor was too nar- 

40 This section is based on Gillespie, The Pacific, 
pp. 142—59; Morison, Breaking the Bismarchs Bar- 
rier, pp. 293-96; Rentz, Bougainville and the 
Northern Solomons, pp. 92-114; ONI USN, Combat 
Narratives: Solomon Islands Campaign, XII, The 
Bougainville Landing and the Battle of Empress 
Augusta Bay, 27 October— 2 November 1943 [Wash- 
ington, 1945], 11-23; SOPACBACOM, The Bou- 
gainville Campaign, Ch. II, OCMH; 8th Area Army 
Operations, Japanese Monogr No. 110 (OCMH), p. 
102; 17th Army Operations, II, Japanese Monogr 
No. 40 (OCMH), 100-103; Outline of Southeast 
Area Naval Air Operations, Ft. IV, Japanese Monogr 
No. 108 (OCMH), p. 44. 

" In Samuel Eliot Morison's words, "The his- 
torian wishes that the exploring captains of H.M.S. 
Blanche, Renard, and Gazelle had not been so 
fond of their ships as to name several harbors, 
channels, and sounds after each one." Breaking the 
Bismarchs Barrier, p. 293, n. 1. 



row to permit ships to maneuver safely, 
the fire support destroyers and seven 
APD's remained west of the harbor. 
While the troops boarded the landing 
craft, destroyers opened fire on the land- 
ing beaches on Mono's south shore, and 
the minesweepers checked Blanche Har- 
bor. At the same time the APD McKean 
put a radar party ashore on Mono's 
north coast. 

Covered by the destroyers' gunfire 
and accompanied by the LCI gunboats, 
the first wave of LCP(R)'s, carrying ele- 
ments of two battalions of the 8th Bri- 
gade, moved through the channel in 
the wet, misty half-light. There were 
only a handful of Japanese on Mono, 
some 225 men of the special naval land- 
ing forces. The naval bombardment 
drove most of the defenders out of their 
beach positions, and as the New Zealand 
infantry went ashore they drove out or 
killed the Japanese in the vicinity of the 
beach. However, enemy mortars and ma- 
chine guns from hidden positions in the 
jungle fired on the landing beaches and 
on the LST's of the fourth transport 
group, which beached at 0735. This fire 
caused some casualties, damaged some 
weapons and equipment, and delayed 
the unloading. But before noon the 8th 
Brigade troops captured two 75-mm. 
guns and one 90-mrn. mortar and resist- 
ance to the landing ceased. 

Stirling Island, which was not occu- 
pied by the enemy, was secured by a 
battalion during the morning. A total of 
2,500 men— 252 Americans of the 198th 
Coast Artillery and several detachments 
from other units, the rest New Zealand- 
ers— had been landed on the south shore 
of Mono. The radar detachment and ac- 
companying combat troops that had 

landed on the north coast of Mono num- 
bered 200. 

Meanwhile the American destroyers 
were busy. In addition to providing fire 
support for the landings they escorted 
the unloaded transport groups back to 
Guadalcanal. Two picket destroyers with 
fighter director teams aboard were sta- 
tioned east and west of the Treasuries 
to warn against enemy air attacks. 

General Hyakutake had decided that 
the Treasury landings were a prelim- 
inary to a systematic operation, and that 
the Allies would build an airfield on the 
Treasuries, take Choiseul, and after 
intensified air and surface operations, 
would land three divisions on southern 
Bougainville in late November. He felt 
that they might possibly invade Buka. 
Warning that the recent decline in Jap- 
anese naval strength might cause the 
Allies to move faster, he stressed the 
importance of building up the south 
Bougainville defenses. In short, he be- 
lieved just what the Allies hoped he 

When Admiral Kusaka at Rabaul was 
notified of the Allied landing, he 
brought some planes forward from Kavi- 
eng and sent fighters and dive bombers 
against the Allies. Most of these were 
headed off by the New Georgia-based 
P-38's and P-40's that formed the south- 
ern force's air cover, but some got 
through to damage the picket destroyer 
Cony and harass the retiring LST's. The 
Japanese pilots reported that they had 
sunk two transports and two cruisers. 

On shore, Brigadier R. A. Row of the 
New Zealand Army, the landing force 
commander, set up beach defenses. By 
12 November his troops had killed or 
captured the enemy garrison which had 



fled into the hills of Mono. Two hundred 
and five Japanese corpses were counted; 
40 New Zealanders and 12 Americans 
had been killed, 145 New Zealanders 
and 29 Americans wounded. 

Succeeding transport echelons, thir- 
teen in all, brought in more troops and 
equipment from 1 November 1943 
through 15 January 1944. During this 
period the boat pool, an advanced naval 
base, and radars were established; these 
supported the main operation at Em- 
press Augusta Bay. Seabees of the U.S. 
Navy built a 5,fioo-foot-long airstrip on 
Stirling that was ready to receive fighter 
planes on Christmas Day. 

The Choiseul Raid 

Four of the APD's that had carried 
Brigadier Row's troops to the Treasur- 
ies sailed to Vella Lavella on 27 Octo- 
ber and there took aboard 725 men of 
Lt. Col. Victor H. Krulak's 2d Marine 
Parachute Battalion, plus fourteen days' 
rations and two units of fire. Escorted by 
the destroyer Conway, the APD's steamed 
for the village of Voza on Choiseul, and 
that night landed the parachutists and 
their gear. 

General Vandegrift had ordered Kru- 
lak so to conduct operations that the 
Japanese would believe a large force 
was present. Krulak therefore raided 
a barge staging point at Sagigai, some 
eight miles from Voza, and then sent 
strong combat patrols to the western 
part of Choiseul. But by 2 November 
the Japanese appeared to be concentrat- 
ing at Sagigai with the obvious intention 
of destroying the 2d Parachute Battal- 
ion. From eight hundred to one thou- 
sand enemy were reported to have moved 
into Sagigai from positions farther east, 

with more on the way. By now the Em- 
press Augusta Bay landing had been 
safely executed, and Vandegrift ordered 
Krulak to withdraw. The battalion em- 
barked on three LCI's in the early morn- 
ing hours of 4 November. The raid cost 
11 Marines dead, 14 wounded; 143 Jap- 
anese were estimated to have been slain. 

Japanese sources do not indicate 
what estimates Imamura and Hyakutake 
placed on the operation. However, since 
Hyakutake expected that Choiseul 
would be invaded after the Treasuries 
and before southern Bougainville, it is 
not unlikely that Krulak's diversion con- 
firmed his belief that southern Bougain- 
ville was the main Allied objective. 

Seizure of Empress Augusta Bay 

Supporting Operations 

In invading Empress Augusta Bay, 
Halsey's forces were bypassing formi- 
dable enemy positions in southern Bou- 
gainville and the Shortlands, and plac- 
ing themselves within close range of all 
the other Bougainville bases, as well as 
within fighter range of Rabaul— thus the 
strong air attacks by the Fifth Air Force 
and the Air Command, Solomons. In 
addition, Halsey had planned to make 
sure that the Japanese bases on Bou- 
gainville were in no condition to launch 
air attacks during the main landings on 
1 November. 12 Forces assigned to this 

"This subsection is based on Halsey and Bryan, 
Admiral Halsey's Story, pp. 177-79; Morison, Break- 
ing the Bismarcks Barrier, pp. 292-93; Admiral 
Frederick C. Sherman, Combat Command: The 
American Aircraft Carriers in the Pacific War (New 
York; E. P. Dutton & Co., Inc., 1950), pp. 199-200; 
ONI USN, The Bougainville Landing and the 
Battle of Empress Augusta Bay, pp. 25-37. 



mission were the 2 carriers, 2 antiair- 
craft light cruisers, and 10 destroyers of 
Admiral Sherman's Task Force 38 and 
the 4 light cruisers and 8 destroyers of 
Admiral Merrill's Task Force 39. 

Task Force 38 sortied from Espiritu 
Santo on 2g October, Task Force 39 
from Purvis Bay on Florida Island on 3 1 
October. Both were bound initially for 

Merrill, sailing well south of the Rus- 
sells and west of the Treasuries on his 
537-mile voyage in pursuance of Hal- 
sey's tight schedule, got there first. He 
arrived off Buka Passage at 0021, 1 No- 
vember, and fired 300 6-inch and 2,400 
5-inch shells at Buka and Bonis fields. 
Shore batteries replied but without ef- 
fect. Merrill then retired at thirty knots 
toward the Shortland Islands. Enemy 
planes harassed the task force but the 
only damage they did was to the ad- 
miral's typewriter. One fire started by 
the bombardment was visible from sixty 
miles away. 

About four hours after the beginning 
of Merrill's bombardment Task Force 
38 reached a launching position some 
sixty-five miles southeast of Buka. This 
was the first time since the outbreak 
of the war in the Pacific that an Allied 
aircraft carrier had ventured within 
fighter range of Rabaul, and the first 
tactical employment of an Allied carrier 
in the South Pacific since the desperate 
battles of the Guadalcanal Campaign. 
In Admiral Sherman's words: 

"We on the carriers had begun to 
think we would never get any action. All 
the previous assignments had gone to 
the shore-based air. Admiral Halsey had 
told me that he had to hold us for use 

against the Japanese fleet in case it came 
down from Truk. . . ." 43 

The weather was bad for carrier op- 
erations as the planes detailed for the 
first strike, a force made up of eighteen 
fighters, fifteen dive bombers, and eleven 
torpedo bombers, prepared to take off in 
the darkness. The sea was glassy and 
calm; occasional rain squalls fell. There 
was no breeze blowing over the flight 
decks, and the planes had to be cata- 
pulted into the air, a slow process that, 
coupled with the planes' difficulties in 
forming up in the dark, delayed their ar- 
rival over Buka until daylight. Two tor- 
pedo bombers and one dive bomber hit 
the water upon take-off, doubtless be- 
cause of the calm air. The rest of the 
planes dropped three 1 ,000-pound bombs 
on Buka's runway and seventy-two 100- 
pound bombs on supply dumps and dis- 
persal areas. 

The next strike— fourteen fighters, 
twenty-one dive bombers, and eleven tor- 
pedo bombers— was launched at 0930 
without casualties. These planes struck 
Buka again and bombed several small 
ships offshore. At dawn the next morn- 
ing, 2 November, forty-four planes at- 
tacked Bonis, and at 1036 forty-one more 
repeated the attack. Then Sherman, 
under orders from Halsey, headed for 
the vicinity of Rennell, due south of 
Guadalcanal, to refuel. In two days of 
action Task Force 38, operating within 
sixty-five miles of Buka, estimated that 
it had destroyed about thirty Japanese 
planes and hit several small ships. More 
important, it had guaranteed that the 
Buka and Bonis runways could not be 
used for air attacks against Admiral Wil- 

Sherman, Combat Command, p. 200. 



Maj. Gen. Allen H. Turnage (right) and Commodore Laurence F. Reifsnider 
aboard a transport before the landings on Bougainville. 

kinson's ships. The Americans lost seven 
men and eleven planes in combat and 
operational crashes. 

Meanwhile Merrill's ships had sped 
from Buka to the Shortlands in the early 
morning hours of 1 November to bom- 
bard Poporang, Ballale, Faisi, and 
smaller islands. Merrill had bombarded 
these before, on the night of 29-30 June, 
but in stormy darkness. Now the bom- 
bardment was in broad daylight; it 
started at 0631, seventeen minutes after 
sunrise. Japanese shore batteries replied 
with inaccurate fire. Only the destroyer 
Dyson was hit, and its casualties and 

damage were minor. His mission com- 
pleted, Merrill headed south. 

Approach to the Target 

The last days of October found Wil- 
kinson's ships busy loading and rehears- 
ing at Guadalcanal and the New Hebri- 
des. 44 Wilkinson had organized his eight 

"The rest of this chapter is based on Morison, 
Breaking the Bismarcks Barrier, pp. 296-305; Rentz, 
Bougainville and the Northern Solomons, pp. 21— 
39; ONI USN, The Bougainville Landing and the 
Battle of Empress Augusta Bay, pp. 37^19; SOPAC- 
BACOM, The Bougainville Campaign, Ch. HI, 
OCMH; 8th Area Army Operations, Japanese 
Monogr No. 110 (OCMH), pp. 103-05; Southeast 



transport and four cargo ships of Task 
Force 31's northern force into three 
transport divisions of four ships each. A 
reinforced regiment of marines was to be 
carried in each of two of the divisions, 
the reinforced 3d Marine Defense Bat- 
talion in the third. The four transports 
of Division A, carrying 6,421 men of the 
3d Marines, reinforced, departed Es- 
piritu Santo on 29 October and steamed 
for Koli Point on Guadalcanal. There 
Admiral Wilkinson and General Vande- 
grift boarded the George Clymer. Gen- 
eral Turnage, 3d Marine Division com- 
mander, and Commodore Laurence F. 
Reifsnider, the transport group com- 
mander, had come up from the New 
Hebrides rehearsal in the Hunter Lig- 
gett. Transport Division B, after the re- 
hearsal, took the 6,103 men °f the rein- 
forced 9th Marines from the New Hebri- 
des and in the late afternoon of 30 Oc- 
tober joined with the four cargo ships 
of Transport Division C south of San 
Cristobal. Division C carried the rein- 
forced 3d Marine Defense Battalion, 
1,400 men, and a good deal of heavy 

All transport divisions, plus 1 1 destroy- 
ers, 4 destroyer-minesweepers, 4 small 
minesweepers, 7 minelayers, and 2 tugs, 
rendezvoused in the Solomon Sea west 
of Guadalcanal at 0740, 31 October. 
They sailed northwestward until 1800, 
then feinted toward the Shortlands, and 
after dark changed course again toward 
the northwest. During the night run to 
Empress Augusta Bay PB4Y4's (Navy 

Area Naval Operations, III, Japanese Monogr No. 
50 (OCMH), 11-13; Outline of Southeast Area 
Naval Air Operations, Pt. IV, Japanese Monogr No. 
108 (OCMH), p. 46; observations of the author, 
who participated as a member of D Battery, 3d 
Marine Defense Battalion. 

Liberators), PV-i's (Vega Ventura night 
fighters), and PBY's (Black Cats) covered 
the ships. Enemy planes were out that 
night and made contact with the cover- 
ing planes but apparently did not spot 
the ships, for none was attacked and Jap- 
anese higher headquarters received no 

Empress Augusta Bay was imperfectly 
charted and the presence of several un- 
charted shoals was rightly suspected. 
Consequently Wilkinson delayed arrival 
at the transport area until daylight so 
that masthead navigation could be used 
to avoid the shoals. 

The Landings 

At 0432 of 1 November, Wilkinson's 
ships changed course from northwest to 
northeast and approached Cape Toro- 
kina in Empress Augusta Bay. Speed was 
reduced from fifteen to twelve knots. 
The minesweepers went out ahead to 
check the area. General quarters sounded 
on all ships at 0500, and forty-five 
minutes later the ships reached the trans- 
port area. The transport Crescent City 
struck a reef but suffered no damage. 

Sunrise did not come until 0614, but 
the morning was bright and clear enough 
for the warships to begin a slow, deliber- 
ate bombardment of Cape Torokina at 
0547. As each transport passed the cape 
it too fired with its 3-inch and antiair- 
craft guns. Wilkinson set H Hour for 
the landing at 0730. At 0645 the eight 
transports anchored in a line pointing 
north-northwest about three thousand 
yards from shore; the cargo ships formed 
a similar line about five hundred yards 
to seaward of the transports. 

Wilkinson, sure that the Japanese 
would launch heavy air attacks, had come 



Mount Bagana 

so lightly loaded that four to five hours 
of unloading time would find his ships 
emptied. Vandegrift and Turnage, an- 
ticipating little opposition at the beach, 
had planned to speed unloading by send- 
ing more than seven thousand men 
ashore in the assault wave. They would 
land along beaches (eleven on the main- 
land and one on Puruata Island off Cape 
Torokina) with a total length of eight 
thousand yards. 

The assault wave boarded landing 
craft at the ships' rails. The winchmen 
quickly lowered the craft into the water; 
and the first wave formed rapidly and 
started for shore. 

The scene was one to be remembered, 
with torpedo bombers roaring overhead, 

trim gray destroyers firing at the beaches, 
the two lines of transports and cargo 
ships swinging on their anchors, and 
the landing craft full of marines churn- 
ing toward the enemy. This scene was 
laid against a natural backdrop of awe- 
some beauty. The early morning tropical 
sun shone in a bright blue sky. A 
moderately heavy sea was running, so 
that at the shore a white line showed 
where the surf pounded on the black and 
gray beaches, which were fringed for 
most of their length by the forbidding 
green of the jungle. Behind were the 
rugged hills, and Mount Bagana, tower- 
ing skyward, emitting perpetual clouds 
of smoke and steam, dominated the en- 
tire scene. 



The destroyers continued firing until 
0731, when thirty-one torpedo bombers 
from New Georgia bombed and strafed 
the shore line for five minutes. The first 
troops reached the beach at 0726, and in 
the next few minutes all the assault wave 
came ashore. There was no opposition 
except at Puruata Island and at Cape 
Torokina and its immediate vicinity. 
There the Japanese, though few in num- 
bers, fought with skill and ferocity. 

Cape Torokina was held by 270 Jap- 
anese soldiers of the 2d Company, 1st 
Battalion, and of the Regimental Gun 
Company, 23d Infantry. One platoon 
held Puruata. On Cape Torokina the 
enemy had built about eighteen log-and- 
sandbag pillboxes, each with two ma- 
chine guns, mutually supporting, camou- 
flaged, and arranged in depth. He had 
also emplaced a 75-mm. gun in an open- 
ended log-and-sand bunker to fire on 
landing craft nearing the beach. 

Neither air bombardment nor naval 
gunfire had had any appreciable effect 
on these positions. Because air recon- 
naissance had shown that the enemy had 
built defense positions on Cape Toro- 
kina (a low, flat, sandy area covered with 
palm trees), it had been a target for 
naval bombardment. Two destroyers had 
fired at the cape from the south, but 
had done no damage. Exploding shells 
and bombs sent up smoke and dust that 
made observation difficult; some shells 
had burst prematurely in the palm trees. 
Poor gunnery was also a factor, for many 
shells were seen to hit the water. 45 

Thus when landing craft bearing the 
3d Marines neared the cape the 75-mm. 

* See Rentz, Bougainville and the Northern 
Solomons, p. 34; Isely and Crowl, The US. Marines 
and Amphibious War, p. 180. 

gun and the machine guns opened fire. 
The men were forced to disembark 
under fire and to start fighting the mo- 
ment they put foot to the ground. Casual- 
ties were lighter than might have been 
expected— 78 men were killed and 104 
wounded in the day's action— but only 
after fierce fighting and much valor were 
the men of the 3d Marines able to estab- 
lish themselves ashore. The pillboxes 
were reduced by three-man fire teams: 
one BAR man and two riflemen with 
M 1 's, all three using grenades whenever 
possible. The gun position was taken by 
Sgt. Robert A. Owens of A Company, 
3d Marines, who rushed the position 
under cover of fire from four riflemen. 
He killed part of the Japanese crew and 
drove off the rest before he died of 
wounds received in his assault. 46 

By 1 1 00 Cape Torokina was cleared. 
Most of its defenders were dead; the 
survivors retreated inland. Puruata Is- 
land was secured at about the same time, 
although some Japanese remained alive 
until the next day. Elsewhere the land- 
ing waves, though not opposed by the 
enemy, pushed inland slowly through 
dense jungle and a knee-deep swamp 
that ran two miles inland and backed 
most of the beach north and east of 
Cape Torokina. The swamp's existence 
had not previously been suspected. 

Air Attacks and Unloading 

The Allied air forces of the South and 
Southwest Pacific Areas had performed 
mightily in their effort to neutralize 
the Japanese air bases at Rabaul, Bou- 
gainville, and the Shortlands, but they 

10 Sergeant Owens received the Medal of Honor 



30 Marines Landing on Cape Torokina 

had not been able to neutralize Rabaul 
completely. In planning the invasion of 
Empress Augusta Bay, the South Pacific 
commanders were aware that the Jap- 
anese would probably counterattack 
from the air. General Twining had ar- 
ranged for thirty-two New Georgia-based 
fighter planes of all types then in use 
in the South Pacific— Army Air Forces 
P-38's, New Zealand P-40's, and Marine 
F4U's— to be overhead in the vicinity 
all day. These planes were vectored by 
a fighter director team aboard the de- 
stroyer Conway. Turning in an out- 
standing performance, they destroyed or 
drove off most of the planes that the 
Japanese sent against Wilkinson. But 
they could not keep them all away. 

At 0718, as the last boats of the assault 
nave were leaving their transports, the 
destroyers' radars picked up a flight of 
approaching enemy planes then fifty 
miles distant. The covering fighters kept 
most of the planes away, but a few, per- 
haps twelve, dive bombers broke through 
to attack the ships. 

These bombers had come from Ra- 
baul, where the enemy commanders were 
making haste to organize counterattacks. 
On 30 October Vice Adm. Sentaro 
Omori, commanding a heavy cruiser di- 
vision, had brought a convoy into Simp- 
son Harbor at Rabaul. Next morning a 
search plane reported an Allied convoy 
of three cruisers, ten destroyers, and 
thirty transports near Gatukai in the 



New Georgia group. This was probably 
Merrill's task force; it could not have 
been Wilkinson's. On receiving this re- 
port Admiral Kusaka ordered the planes 
of his nth Air Fleet to start attacks, and 
he and Koga, over the protests of the 
8th Fleet commander, who warned of the 
dangers of sending surface ships south 
of New Britain, directed Omori to take 
his force and all the 8th Fleet ships out 
to attack. This Omori did, but he missed 
Merrill and returned to Rabaul on the 
morning of l November. 

Then came the news of the landing at 
Empress Augusta Bay. General Hyaku- 
take was still sure that the main Allied 
attack would be delivered against south- 
ern Bougainville, but General Imamura 
ordered him to destroy the forces that 
had landed. Imamura also arranged with 
Kusaka for a counterattacking force from 
the iyth Division, made up of the 2d 
Battalion, 54th Infantry, and the 6lh 
Company, 2d Battalion, 53d Infantry, to 
be transported to Empress Augusta 
Bay. 47 It would be carried on 6 destroyer- 
transports and escorted by 2 heavy cruis- 
ers, 2 light cruisers, and 6 destroyers, all 
under Omori. 

Admirals Koga and Kusaka, just com- 
pleting their preparations for Operation 
RO, also ordered out their planes. The 
weather had come to their assistance by 
halting the heavy raid General Kenney 
had planned for 1 November. Koga alert- 
ed the 12th Air Fleet for transfer from 
Japan to Rabaul. Kusaka sent out planes 
of his nth Air Fleet. The carrier planes 

" This force had been standing by awaiting orders 
to move to western New Britain. It was separate 
from the ijth Division units scheduled for Bou- 
gainville mentioned above, p. 238. 

apparently did not take part on 1 No- 

According to enemy accounts, Japa- 
nese planes delivered three separate at- 
tacks against Wilkinson on 1 November. 
The Japanese used a total of 16 dive 
bombers and 104 fighters, of which 19 
were lost and 10 were damaged. 48 

When Wilkinson's ships received 
warning at 0718, the transports and cargo 
ships weighed anchor and steamed for 
the open sea. They escaped harm, and 
the dive bombers were able to inflict 
only light damage to the destroyers. Two 
sailors were killed. The transports re- 
turned and resumed unloading at 0930, 
having lost two hours. 

Another enemy attack at 1248 suc- 
ceeded in breaking through the fighter 
cover. Warned again by radar, the trans- 
ports, with the exception of the Amer- 
ican Legion, which stuck on an unchart- 
ed shoal, fled. The Japanese attacked the 
moving ships instead of the Legion. No 
damage was done, but the ships lost two 
more hours of unloading time. 49 

The halts in unloading caused by air 
attacks, coupled with beach and terrain 
conditions that Admiral Halsey de- 
scribed as "worse than any we had pre- 
viously encountered," slowed the move- 
ment of supplies and equipment. 30 Fully 
one third of the landing force— 5,700 
men in all— had been assigned to the 
shore party, but nature and the Japanese 

"Southeast Area Naval Operations, III, Japanese 
Monogr No. go (OCMH), p. 46; p. 13 states that 
twenty-two planes were lost. 

" As usual the Japanese claims, like those of 
American pilots, were exaggerated. They said they 
sank two transports and a cruiser. 

50 Halsey and Bryan, Admiral Halsey's Story, p. 

aircraft thwarted efforts to unload all 
the ships on D Day. 

Even on quiet days the surf at Em- 
press Augusta Bay was rough, and on 1 
November a stiff breeze whipped it high- 
er. The northernmost beaches were steep 
and narrow. The surf, and possibly the 
inexperience of some of the crews, took 
a heavy toll of landing craft. No less 
than sixty-four LCVP's and twenty-two 
LCM's broached on shore and were 
swamped by the driving surf. As surf 
conditions got worse, several beaches be- 
came completely unusable. Five ships 
were shifted to beaches farther south, 
with more delay and congestion at the 
southern beaches. It was during this 
move that the American Legion ran 

By 1730 the eight transports were 
empty and Wilkinson took them back to 
Guadalcanal. But the four cargo ships, 
which carried heavy guns and equip- 
ment, were still practically full. Vande- 
grift, who had had ample experience at 
Guadalcanal in being left stranded on a 
hostile shore while much of his equip- 
ment remained in the holds of depart- 
ing ships, persuaded Wilkinson to allow 
the cargo ships to put out to sea for the 
night and return the next morning to 
unload. 51 Most of the troops aboard went 
ashore in LCVP's before Commodore 
Reifsnider led the cargo ships out to sea. 
D Battery of the 3d Defense Battalion, 
for example, its 90-mrn, antiaircraft 

51 Miller, Guadalcanal: The First Offensive, pp. 



guns, fire control equipment, and radars 
deep in the holds of the Alchiba, which 
had lost all its LCM's in the raging surf, 
went ashore as infantry and occupied a 
support position in the sector of the gth 

Except for the full holds of the cargo 
ships, D Day had been thoroughly suc- 
cessful. All the landing force, including 
General Turnage, Brig. Gen. Alfred H. 
Noble, corps deputy commander, Col. 
Gerald C. Thomas, the corps chief of 
staff, and several other officers, were 
ashore. General Vandegrift returned to 
Guadalcanal on the George Glymer, leav- 

ing Turnage in command at Cape Toro- 
kina. By the day's end the division held 
a shallow beachhead from Torokina 
northward for about four thousand 
yards. Aside from unloading the cargo 
ships (a task that was expeditiously ac- 
complished the next day), the main mis- 
sions facing the amphibious and ground 
commanders and the troops were three- 
fold: to bring in reinforcements; to or- 
ganize a perimeter defense capable of 
beating off the inevitable Japanese coun- 
terattack; and to build the airfields that 
would put South Pacific fighter planes 
over Rabaul. 


Exploiting the Beachhead 

The ground troops at Cape Torokina 
could be expected to carry out their mis- 
sions efficiently only if they were unham- 
pered by Japanese aircraft and warships. 
Therefore the real battle for the beach- 
head was fought in the air and on the 
sea. The primary mission of South Pa- 
cific aircraft and warships during the 
first days of November was protection 
of the newly won beachhead. In this 
mission they fought hard and with excel- 
lent results. 

Air and Surface Action, i-ii November 

When Admiral Omori led his task 
force out of Rabaul in late afternoon of 
1 November, he had orders to escort 
Imamura's troops and to attack Wilkin- 
son's transports in Empress Augusta 
Bay. But after joining with the troop- 
carrying destroyers in Saint George's 
Channel between New Britain and New 
Ireland, Omori was sighted by a U.S. 
submarine. Further, an unidentified 
plane dropped a bomb near the light 
cruiser Sendai. The Japanese, sure that 
their intentions had been deduced, post- 
poned the troop movement, but Omori 
was allowed to take his task force of 
two heavy and two light cruisers and 
six destroyers to Empress Augusta Bay 
with the intention of destroying the 

American transports and cargo ships 
which he thought would still be there. 1 

Meanwhile, Admiral Merrill's Task 
Force 39 had sailed to the vicinity of 
Vella Lavella after the two bombard- 
ments on 1 November. Four of his 
eight destroyers were refueling in the 
late afternoon of 1 November when 
General Twining's reconnaissance planes 
spotted Omori and flashed a warning. 
Halsey ordered Merrill out to intercept 
Omori. Receiving continuous, accurate 
plots of Omori's course and speed, Mer- 
rill set his course and speed so that his 
four light cruisers and eight destroyers 
would intercept west of Empress Au- 
gusta Bay. 

At 0229, 2 November, a few miles 
from Cape Torokina, Task Force 3g 
made contact with Omori and attacked 
at once. In this engagement, the Battle 
of Empress Augusta Bay, Merrill sank 
one light cruiser and one destroyer; ex- 
cept for the destroyer Foote, which lost 
her stern to a Japanese torpedo, the 

1 This section is based on Craven and Cate, The 
Pacific: Guadalcanal to Saipan, pp. 257-61, 325—28; 
Halsey and Bryan, Admiral Halsey's Story, pp. 180- 
85; Morison, breaking the Bismarcks Barrier, pp. 
305-37; Sherman, Combat Command, 201—08; CTF 
33 War Diary; Southeast Area Naval Operations, 

III, Japanese Monogr No. 50 (OCMH), 14-29; Out- 
line of Southeast Area Naval Air Operations, Pt. 

IV, Japanese Monogr No. 108 (OCMH), pp. 48-68. 



American ships received light damage. 
The flashes from gunfire and explosions 
were visible to Commodore Reifsnider's 
four cargo ships, which had put out to 
sea, and to the marines ashore. The en- 
gagement lasted until dawn, when 
Omori, tacitly acknowledging failure, 
took his surviving ships back to Rabaul. 

Near as he was to Rabaul, Merrill ex- 
pected to suffer air attack at dawn, and 
he was not wrong. When a Japanese 
patrol plane sighted him 18 dive bomb- 
ers and 80 fighters promptly took off 
from Rabaul to the attack. Bad weather 
on the morning of 2 November had kept 
most of the Allied fighters on the New 
Georgia fields, but 8 F6F's, 1 F4U, 3 
P-38's, and 4 New Zealand P-40's, vec- 
tored by a destroyer still in Empress 
Augusta Bay, hurled themselves at part 
of the Japanese formation and shot down 
several planes. 

The remaining enemy planes came 
upon Task Force 39 shortly before 0800 
and promptly attacked. The task force 
maneuvered rapidly, sailing clockwise in 
a great circle and shooting 5-inch, 40- 
mm., 20-mm., and even 6-inch guns at 
the diving Japanese with considerable 
success. The light cruiser Montpelier 
suffered two bomb hits which wounded 
several men, but the other ships went 
unscathed. The Japanese broke off the 
action, but on the way home lost more 
planes to Allied fighters. More planes 
from Rabaul would doubtless have come 
out after Merrill that day but for the 
Fifth Air Force's raid on the airfields, 
which the Japanese carrier pilots con- 
tested so hotly. 

Merrill's ships, after two busy days 
that included two shore bombardments, 
the night action of Empress Augusta 

Bay, and the morning air attack, now 
escorted Reifsnider's retiring cargo ships 
as far as Rendova, then steamed for 
Florida and concluded their eventful, 
successful cruise. On the other hand, the 
Japanese had lost two ships and numer- 
ous aircraft, and had not inflicted any- 
thing like equivalent damage to the 
Americans. But Admiral Koga had not 
given up. When he was informed of the 
landings at Empress Augusta Bay, he 
ordered Vice Adm. Takeo Kurita to take 
seven heavy cruisers, one light cruiser, 
four destroyers, and a fleet train from 
Truk to Rabaul. Kurita arrived safely 
on 4 November, although later ships 
were hit by Twining's B-24's. 

This force of heavy cruisers at Rabaul 
posed a serious threat to the new beach- 
head at Empress Augusta Bay, and cre- 
ated, wrote Admiral Halsey, "the most 
desperate emergency that confronted me 
in my entire term as COMSOPAC." 2 
He knew that he had to stop them, but 
he had only two naval task forces- 
Merrill's, which was exhausted after its 
performance of 1-2 November, and 
Sherman's carriers. Up to now carriers 
had been employed against land bases 
only in the most gingerly fashion. The 
South Pacific staff calculated that Sher- 
man, from his refueling position near 
Rennell, could strike Kurita before 
Kurita would strike Empress Augusta 
Bay. So Halsey ordered Sherman to hit 
Rabaul. When he gave these orders the 
South Pacific commander expected the 
carrier air groups to be "cut to pieces" 
and the carriers "stricken." 3 

"I fully expected that they [Sherman's 

2 Halsey and Bryan, Admiral Halsey's Story, pp. 
'Ibid., p. 181. 



carriers] would be lost." 4 ". . . but we 
could not let the men at Torokina be 
wiped out while we stood by and wrung 
our hands." 5 Halsey was never a man 
to stand idly by and wring his hands, or 
to allow anyone else that emotional 

Halsey directed South Pacific land- 
based air (Task Force 33) to provide 
cover for Sherman during his daylight 
approach and retirement. This job was 
done by Navy fighters from New Geor- 
gia, which of course were capable of 
landing on carrier decks. Thus Sherman 
was able to send all his aircraft against 
Rabaul instead of keeping some of them 
overhead for protection. 

Task Force 38 reached its launching 
point in the Solomon Sea 57 miles south- 
west of Torokina and 230 miles south- 
east of Rabaul at 0900, 5 November. The 
weather was fine for carrier operations; 
a steady breeze was blowing, and there 
were frequent rain squalls where the 
ships could hide in case of air attack. 
The two carriers sent out 97 planes: 23 
torpedo bombers, 22 dive bombers, and 
52 fighters. They arrived over Rabaul 
and dived through a hole in the clouds 
to take the Japanese by surprise. Though 
faced by intense antiaircraft fire they 
bored in with resounding success. They 
did not sink any ships, but damaged 
three heavy cruisers, two light cruisers, 
and two destroyers so severely that 
months passed before any of them were 
fit to fisfht aofain. This was done at a 
cost of fifteen men killed or missing, ten 

1 Halsey's preface to Sherman's Combat Com- 
mand, p. 8. 

5 Halsey and Bryan, Admiral Halsey's Story, p. 
1 Si. 

planes lost. Halsey's gloomy expectations 
were not fulfilled. 

Twenty-seven B-24's and fifty-eight 
P-38's from the Fifth Air Force reached 
Rabaul in the afternoon. As practically 
all the Japanese planes were out after 
Task Force 38, Kenney's men bombed 
the wharves. The Japanese failed to find 
Sherman, but they attacked an LCI, an 
LCT, and a PT boat near the southern 
arm of Empress Augusta Bay, and 
claimed a tremendous but nonexistent 

Sherman's victory, on the other hand, 
was real. Next day Koga decided to pull 
his heavy cruisers back to Truk, and the 
threat to Cape Torokina was ended. 
Thereafter no more heavy Japanese ships 
went to Rabaul. 

Meanwhile Kusaka's nth Air Fleet 
and the carrier planes, besides attacking 
Merrill and Sherman, had been striking 
day and night against Cape Torokina, 
hammering at reinforcement convoys, 
and fighting almost constantly with Al- 
lied fighter planes. They damaged three 
ships and sank one, but kept losing 
planes to ship- and shore-based antiair- 
craft guns and to Twining's fighters. 

Air Command, Solomons, made a max- 
imum effort to keep the enemy's Bou- 
gainville bases out of action and to keep 
the Rabaul-based planes away from Cape 
Torokina and the reinforcement con- 
voys. For example, on 10 November 
there were 7 1 2 take-offs and landings 
at Munda airfield alone. 

Rabaul was still a primary target for 
General Kenney. The weather prevented 
an attack on 6 November, but 10 No- 
vember saw a heavy attack, and next day 

"For details see Morison, Breaking the Bismarcks 
Barrier, p. 329. 



Aircrewman Wounded in Strike on Rabaul is helped out of his plane on flight 
deck of aircraft carrier USS Saratoga, 5 November 1943. 

RAAF Beauforts and Fifth Air Force 
planes struck in the morning before 
heavy clouds piled up over Rabaul. 

The additional carrier task group of 
the Fifth Fleet that Admiral Nimitz 
had promised to Halsey reached the 
South Pacific on 7 November. Com- 
manded by Rear Adm. Alfred L. Mont- 
gomery, it consisted of the carriers Essex, 
Bunker Hill, and Independence. Halsey 
planned to use Montgomery's ships as 
well as Task Force 38 in a double carrier 
strike against Rabaul on 11 November. 

Sherman sailed to a point in the Pa- 
cific Ocean near Green Island, north- 
northwest of Bougainville, and launched 
planes. They reached Rabaul in bad 

weather about 0830, struck at ships, and 
returned to the carriers, which retired 
southward without being detected. 

Montgomery launched his strike from 
a point in the Solomon Sea about 160 
miles southeast of Rabaul. His planes 
hit at ships too, then returned to their 
mother carriers. The Japanese found 
Montgomery and delivered a series of 
furious though unsuccessful air attacks 
which inflicted only slight damage. They 
lost thirty-five planes to ships' antiair- 
craft guns and to Allied fighters from 
New Georgia. 

In eleven days of the RO operations 
against the Allied lines of communica- 
tion and the Torokina beachhead, the 



Japanese pilots had reported enormous 
damage to Allied ships and planes, 
whereas in reality they had accomplished 
very little and had suffered the real dam- 
age themselves. 7 Koga had sent 173 
planes and 192 men down from Truk, 
and by the end of 11 November 121 
planes had been destroyed and 86 of the 
men were dead. The nth Air Fleet had 
lost about 70 planes. These losses 
". . . had put the carrier air force in a 
position where further combat would 
rob it of even a skeleton force around 
which to rebuild. . . ." 8 Koga may have 
believed his pilots' claims, but he also 
recognized the significance of his own 
losses. On 12 November he withdrew 
the carrier planes to Truk. The with- 
drawal, first of the cruisers and then of 
the planes, ended Rabaul's offensive 
threat. Thereafter it was a formidable 
defense position only, and after Armis- 
tice Day Southwest Pacific planes were 
able to cease their attacks against it and 
concentrate against enemy bases to the 

The damage that Sherman's pilots in- 
flicted on the heavy cruisers and Koga's 
losses in carrier planes had repercussions 
that were felt far beyond Empress Au- 
gusta Bay. Koga had planned to use the 
Combined Fleet to seek out and destroy 
the U.S. Pacific Fleet if the Americans 
invaded the Gilberts or Marshalls, but 
when Admiral Nimitz' forces moved into 
the Gilberts on 21 November 1943, Koga 
did not stir out of Truk; the cruiser 
damage and aircraft losses had com- 
pletely immobilized the Combined 

Fleet. 9 This series of events, wherein the 
Japanese shifted forces back and forth 
to meet Allied threats from different 
parts of the Pacific, and lost as a result, 
was an advantage the Joint Chiefs of 
Staff had in mind when they ordered 
two advances rather than one. The series 
illustrates also the strategic importance 
of Rabaul, and the advantages that their 
interior lines gave to the Japanese. 

Operations Ashore 

Now landed and completely protected 
from Japanese surface attack, although 
subject to frequent air raids by day and 
by night, the 3d Marine Division was 
hampered as much by terrain as by the 
enemy. The swamps and dense forest 
slowed the movement of supplies and the 
building of roads and airfields. During 
their first five days on shore the marines 
patrolled, established antiaircraft and 
beach defenses, and extended the perim- 
eter two thousand yards inland. (Map 
16) Seventy-eight marines were killed or 
missing, 104 wounded. 10 

More Troops 

The first reinforcements, one battalion 
of the 21st Marine Regiment, arrived on 

7 The Japanese reported sinking 5 battleships, 10 
aircraft carriers, 19 cruisers, 7 destroyers, and 9 
transports between 27 October and 10 December. 

'Southeast Area Naval Operations, III, Japanese 
Monogr No. 50 (OCMH), 25—26. 

6 See Wilds, "The Admiral Who Lost His Fleet," 
United States Naval Institute Proceedings, Vol. 77, 
No. 11 (November 1951); Crowl and Love, The 
Seizure of the Gilberts and Marshalls, Ch. IV. 

"This section is based on Morison, Breaking the 
Bismarcks Barrier, pp, 337-69; Rentz, Bougainville 
and the Northern Solomons, pp. 39—80; [British] 
Central Office of Information, Among Those Pres- 
ent, pp. 64-73; SOPACBACOM, The Bougainville 
Campaign, Chs. III-IV, V, OCMH; Halsey, Narra- 
tive Account of the South Pacific Campaign, 
OCMH; Harmon, The Army in the South Pacific, 
OCMH; I Mar Amphib Corps, Bougainville Beach- 
head; 8 th Area Army Operations, Japanese Monogr 
No. 110 (OCMH), pp. 103-06; 17th Army Opera- 
tions, II, Japanese Monogr No. 40 (OCMH), 104; 
the author's observations. 



Amphibian Tractors LVT (i) , carrying supplies and ammunition, move inland 
over a muddy trail. 

eight LST's and eight APD's on 6 No- 
vember. Escorted by six destroyers and 
covered by Task Force 39, these ships 
had sailed from Purvis Bay two days 
before. Japanese aircraft harried them 
during the night of 5-6 November but 
did no damage. 

For speedy unloading, the LST car- 
goes had been packed on trailers at Pur- 
vis Bay. But Cape Torokina did not 
boast very many beaches suitable for the 
LST's (which in the South Pacific almost 
never carried tanks). One beach at Pu- 
ruata Island had room for three LST's, 
but using this meant unloading gear at 
Puruata and then transshipping it to the 
mainland. At the beaches east of Cape 
Torokina the LST's grounded offshore. 

Seabees improvised coconut log runways, 
which failed to stand up under the 
strain. The eventual answer to the prob- 
lem lay in steel pontons. 

On 8 November substantial reinforce- 
ments came in, some aboard six of the 
ships that had made the initial invasion 
and then returned to Guadalcanal to 
pick up the 148th Regimental Combat 
Team of the 37th Division. Japanese 
aircraft made the day exciting as the sol- 
diers unloaded and went ashore. Over 
a hundred planes attacked at noon. 
Twenty-eight Allied fighters from New 
Georgia kept many of them off, but some 
got through and damaged the President 
Jackson. Once ashore, the 148th relieved 
the 3d Marines on the left flank, and 



the marine regiment was assigned a po- 
sition in the middle of the inland side 
of the perimeter defense. 

General Geiger, having flown out from 
Washington and relieved General Van- 
degrift as corps commander, arrived at 
Bougainville on the gth. On 13 Novem- 
ber Admiral Wilkinson relinquished his 
control and Geiger became directly re- 
sponsible to Halsey. The amphibious 
commander retained responsibility for 
the transport of troops and supplies to 
the beachhead. 

Other reinforcements from the 37th 
and 3d Marine Divisions came in 
promptly. The 129th Regimental Com- 
bat Team landed on 13 November, and 
was followed six days later by the 145th. 

Except for miscellaneous units and de- 
tachments, this completed the movement 
of General Beightler's veteran division 
to the beachhead. The remaining units 
of the 2 1st Marines arrived on 11 and 17 
November. During the latter shipment 
the APD McKean was fatally torpedoed 
by a Japanese plane. Thus by the end 
of the third week in November there 
were two full divisions at Empress 
Augusta Bay, plus substantial bodies of 
corps troops, naval construction battal- 
ions, and naval base forces. The I Marine 
Amphibious Corps held a perimeter 
about sixteen thousand yards in circum- 
ference, including seven thousand yards 
along the beach. The 37th Division held 
the left, the 3d Marine Division the 



right. This perimeter was not attained 
without fighting, but the 37th Division 
was fortunate in that, except for patrol 
clashes, all the fighting occurred in the 
3d Marine Division's zone. 

Expansion of the Perimeter 

Even after 1 November Japanese 
Army commanders continued to cherish 
the delusion that the main effort was yet 
to come, and that southern Bougainville 
or Buka was the real target. However, 
on orders from Imamura to destroy the 
wide, shallow Allied beachhead at Cape 
Torokina, Hyakutake dispatched the two 
battalions of the 23 d Infantry from the 
Buin area to the cape. Under command 
of Maj. Gen. Shun Iwasa, infantry group 
commander of the 6th Division, the 23 d 
was to operate in conjunction with the 
ijth Division troops whose transfer from 
Rabaul, first planned for 1 November, 
had been postponed. Aboard four de- 
stroyers, 475 men of this group finally 
got under way for Torokina on 6 No- 
vember; 700 others sailed for Buka. The 
zyth Division troops were to cover the 
movement of the 23d Infantry by land- 
ing north of the cape near the Ameri- 
cans' left and creating a diversion. They 
would then move inland and join with 
the 23 d. Iwasa was to advance down a 
trail with the combined force and attack 
the beachhead. 

The troop-carrying destroyers hove to 
off the beach between the Laruma and 
Koromokina Rivers in the predawn 
darkness on the morning of 7 November. 
Between 0400 and 0600, the 475 soldiers 
slipped ashore in twenty-one landing 
craft under the very noses of the Ameri- 
can defenders. Patrolling PT boats 
missed the destroyers, and an antitank 

Soldiers of the 148TH Regimental 
Combat Team boarding the transport 
President Jackson for the run to 
Bougainville, 5 November 1943. 



platoon on shore saw the landing craft 
but thought they were American. The 
enemy soldiers landed so close to the 
American lines that they actually cut off 
several marines in an outpost, who were 
later rescued by two LCM's. 

The Japanese attacked at once in the 
vicinity of a lagoon about fourteen hun- 
dred yards west of the Koromokina 
River. The sector was defended by troops 
of the 3d Marines who had just ex- 
changed positions with the gth Marines. 
General Turnage had ordered the trans- 
fer because the 3d had seen all the fight- 
ing on D Day, the gth (which landed on 
the left) none, and there seemed to be 
no immediate prospect of fighting on the 
left. The enemy made some small local 
gains by infiltrating. The fighting, with 
rifles, machine guns, mortars, and gre- 
nades, was close work, but the marine 
lines held. 

Next morning five field artillery bat- 
teries, plus mortars, antitank guns, and 
machine guns, fired a twenty-minute 
preparation into the Japanese position. 
Then the newly arrived 1st Battalion, 
21st Marines, supported by light tanks, 
assaulted. It met only light opposition; 
the artillery preparation had come close 
to achieving perfection. Instead of en- 
gaging in a fierce fight, the 1st Battalion 
walked, cautiously but steadily, through 
the jungle. It found, in the small area 
where the Japanese had packed them- 
selves, about three hundred men killed 
almost instantaneously, their dead bodies 
lying beside their smashed weapons. In 
this action at Koromokina Lagoon the 
marines suffered sixteen men killed, 
thirty wounded. 

Meanwhile the 23d Infantry had 

moved into position inland and had al- 
ready begun attacking the trail blocks 
the marines had set up. Control of the 
trail system inland was of great im- 
portance to the security of the beach- 
head. It was clear that unless the Japa- 
nese had enough strength to deliver a 
major attack from the sea (and Admiral 
Sherman had settled that question on 5 
November) any counteroffensives would 
be delivered along the axes of the trails. 
There were two important tracks at 
Cape Torokina, East-West Trail and the 
Numa Numa Trail. The latter ran from 
the shore near the mouth of the Piva 
River northward through the mountains 
to Numa Numa on the east coast. East- 
West Trail intersected the Numa Numa 
Trail about five thousand yards inland 
(north) of the Piva's mouth. It led east- 
ward, then north through the mountains 
to Roravana Bay and intersected the 
several trails leading to Buin. A local 
track, Mission Trail, ran from a point 
about two thousand yards north of the 
Piva mouth southwestward to the Roman 
Catholic mission station at Buretoni just 
northwest of Torokina. 

On 5 November the 23d Infantry at- 
tacked a block on Mission Trail that was 
held by the 3d Raider Battalion. After 
the raider battalion beat off the 23d; it 
and later the 3d Battalion, gth Marines, 
counterattacked up Mission Trail and by 
Armistice Day had advanced to the junc- 
tion of Mission and Numa Numa Trails. 
Losing ig killed and 32 wounded, the 
marines estimated that they had ac- 
counted for 550 of the enemy. 

Two days later the 21st Marines con- 
tinued the fight, this time not only to 
keep control of the trails but also to se- 



cure an airfield site. Since landing the I 
Marine Amphibious Corps had also been 
hard at work pushing supply routes 
through the swamps, an extremely dif- 
ficult and time-consuming task. At the 
same time patrols had found a good air- 
field site in a coconut grove by the right 
(west) bank of the Piva River near the 
junction of the Numa Numa and East- 
West Trails. This was some distance 
from the 3d Division's front, and the dif- 
ficulties of pushing supplies so far pre- 
vented an immediate forward displace- 
ment of the 3d Division to include the 
site. Generals Geiger and Turnage there- 
fore decided to establish a self-sustaining 
outpost at the trail junction in order to 
hold the airfield site. On 13 and 14 No- 
vember troops of the % 1st Marines, fight- 
ing hard against Japanese in prepared 
positions, made their way through the 
coconut grove and by 1600 of 14 Novem- 
ber had seized the trail junction. 

Because the building of roads and 
trails inside the beachhead eased the 
logistical situation, Geiger decided to 
move his whole front forward in the 
latter part of November. The 3d Di- 
vision would advance on the east (right), 
the 37th Division on the west. Five 
artillery battalions, operating under the 
37th Division artillery commander, Brig. 
Gen. Leo N. Kreber, would provide sup- 
port, as would the Aircraft, Northern 
Solomons, under Brig. Gen. Field Harris, 
USMC. The 37th Division met no fight- 
ing in its advance but the 3d Marine 
Division continued to meet opposition 
from the 23d Infantry along the trails, 
especially on the Numa Numa Trail 
north of the airfield site and in the region 
northeast of that site where the East- 

West Trail crossed several tributary 
forks of the Piva River. Here, between 
20 and 24 November, the Japanese re- 
sisted vigorously but vainly. By 36 No- 
vember the 3d Marine Division, main- 
taining contact on the left with the 37th 
Division, had extended its lines as far 
north as the south shore of Lake Kath- 
leen, about 7,500 yards north of the 
Piva's mouth. In the fighting in the Piva 
forks the 3d Marines took the first high 
ground in the beachhead. Along the 
shore line the I Corps held the beach 
from a point 6,000 yards northwest of 
Cape Torokina to a point 3,500 east of 
the cape. The inland lines of the perim- 
eter were about 19,500 yards long. 

During November the Japanese Army 
commanders still refused to believe that 
Halsey had made his main effort at Em- 
press Augusta Bay and therefore under- 
took no counterattacks on a scale large 
enough to be effective. But Rabaul-based 
aircraft continued to raid the beachhead. 
Both division command posts were hit, 
as were several fuel and ammunition 
dumps, which blew skyward in impres- 
sive and expensive displays. On a few 
occasions the enemy planes swooped 
down suddenly over the mountains dur- 
ing daylight and caught the beachhead 
by surprise (the mountains blocked the 
radar beams), but most of the bombings 
were nocturnal, and the Japanese simpli- 
fied the radar operators' problems by 
attacking from seaward where they were 
easy to locate in time for warning to be 
given and the antiaircraft guns to go 
into action. Puruata Island, with phos- 
phorescent water outlining it clearly, 
was a favorite and profitable target, since 
it was nearly always packed with supplies 

Admiral William F. Halsey, Jr., center, with Maj. Gen. Roy S. Geiger, left, 
and Brig. Gen. Leo N. Kreber, Bougainville, 13 November 1943. 

37™ Division Troops moving inland from the beach over a slimy mud trail, 
8 November 1943. These men are from the 148th Regimental Combat Team. 

Results of Japanese Bombing of Puruata Island, 20 November 1943. Top left, 
view from Bougainville. Note wrecked landing craft, foreground. Top right, 
damaged 90-mm. gun of F Battery, 3d Defense Battalion. Bottom, fuel dump 
on fire. 



awaiting transshipment to the main- 
land. 11 

These attacks did not jeopardize the 
security of the beachhead but they were 
a costly nuisance. Of ninety air alerts in 
November, twenty-two resulted in bomb- 
ings and strafings that killed twenty-four 
men and wounded ninety-six. In addi- 
tion to the antiaircraft guns a few PV-i 
night fighters from New Georgia de- 
fended against Kusaka's fliers. Though 
their losses were lighter than in daylight 
attacks, the Japanese lost several planes 
to the night fighters and the antiaircraft 

So sure were the Japanese that Buka 
was an ultimate target that they con- 
tinued to send reinforcements there. 
Late in November 920 soldiers on board 
three destroyers with two more escort- 
ing attempted to get to Buka. They were 
intercepted in the Solomon Sea during 
the night of 25 November by Captain 
Burke's destroyer squadron, which 
chased them from near Buka almost to 
Cape Saint George, the southern tip of 
New Ireland. Burke's ships sank three 
destroyers without receiving as much as 
one hit themselves. This action, the 
Battle of Cape Saint George, was the last 
of the night surface engagements which 
had characterized the Solomons cam- 
paigns since the one off Savo Island on 
8 August 1942. 12 

In November and December at Em- 
press Augusta Bay the indefatigable Jap- 

11 In the early hours of ao November a Japanese 
plane scored a direct hit on one of the go-mm. guns 
of F Battery, gd Defense Battalion, on Puruata 
that killed live and wounded eight of the crew as 
well as knocking the gun out of commission and 
blowing up a fuel dump. 

" For details see Morison, Breaking the Bismarcks 
Barrier, pp. 352-59. 

anese had begun to emplace artillery of 
calibers as high as 150-mm. on the high 
ground around the beachhead, especially 
in a group of hills that lay east of the 
Numa N uma-East-West Trails' junction 
and paralleled the west bank of the 
Torokina River. With these guns they 
shelled the beachhead, especially the air- 
strips and the supply dumps. The 3d 
Marine Division reacted by extending 
its lines to include the hills in a series 
of operations that lasted from g Decem- 
ber through 27 December. One emi- 
nence, Hellzapoppin Ridge, was a nat- 
ural fortress three hundred feet long, 
with sharp slopes and a narrow crest. It 
overlooked much of the beachhead and 
was an excellent site for artillery. Here 
the Japanese had constructed extensive 
positions on the reverse slopes using nat- 
ural and artificial camouflage. The 21st 
Marines attacked Hellzapoppin Ridge 
but were driven off on 18 December. 
Several air strikes missed the narrow 
ridge completely. Finally, co-ordinated 
air (TBF's dropped 100-pound bombs 
with delay fuzes), artillery, and infantry 
attacks resulted in the capture of Hellza- 
poppin on Christmas Day. In the air 
strikes success was finally attained by 
marking the American front lines with 
colored smoke and designating the enemy 
targets with white phosphorus. 

By 15 December the Americans held 
their final defensive line, a perimeter 
defense that extended on its inland side 
for about 22,500 yards. Over 44,000 men 
were present. Construction of the de- 
fense perimeter had begun in some sec- 
tors on 25 November, and by 15 Decem- 
ber the work was complete. The line 
consisted of two-man foxholes, trenches, 
emplacements for automatic weapons, 



105-MM. Howitzer of the 135th Field 
Artillery Battalion in action. 

mortars, antitank guns, and artillery, 
with alternate positions for all weapons. 
Fields of fire were cleared for 100 yards 
in front of the lines but all possible 
foliage was left in place overhead. The 
field artillery, grouped under command 
of General Kreber, was sited to fire in 
support of any threatened sector, and all 
weapons were registered and adjusted for 
every possible avenue of approach. All 
trails were blocked, and the approaches 
to the swamps were mined. Whenever 
possible machine guns were posted in 
commanding positions on high ground. 
The 4.3-inch chemical mortars were so 
sited and adjusted that they could place 
their fire directly in front of the infantry. 
The whole front was wired in behind 
two rows of either double-apron or con- 
certina barbed wire, and the wire was 

full of trip wires and of cans hung up 
to rattle when an enemy approached the 
wire. Several antiaircraft searchlights 
were set up to illuminate the front at 
night, either directly or by throwing up 
widely spread beams that would be re- 
flected down from the clouds. The de- 
fenses were formidable, and it would be 
some time before the Japanese got 
around to testing them thoroughly. 
Meanwhile life inside the perimeter 
promised to be relatively agreeable. 

The XIV Corps Takes Over 

The 3d Marine Division had borne 
the brunt of operations thus far, but it 
was not to be allowed to settle down in 
comfort behind its defenses. Admiral 
Halsey had other plans. The Americal 
and 40th Divisions had at first been 
scheduled for the projected assault 
against Kavieng, but Halsey now wanted 
the I Marine Amphibious Corps, con- 
sisting of the 3d Marine Division and 
the 40th Infantry Division, to conduct 
the operation. He proposed sending 
General Griswold's XIV Corps head- 
quarters to Bougainville to relieve Gen- 
eral Geiger's headquarters, and trans- 
ferring the Americal Division from the 
Fijis to relieve the 3d Marine Division. 13 
When Halsey first announced his plan 
on 2 November General Harmon op- 
posed it, but Halsey overrode his ob- 

Thus on 15 December General Gris- 
wold relieved General Geiger of com- 
mand of all Allied air, surface, and 
ground forces based at Empress Augusta 
Bay and in the Treasuries. Admiral Hal- 
sey also made Harmon his informal 

"The 40th was then in Hawaii due for shipment 
to the South Pacific. 



deputy for supervising operations of the 
XIV Corps. On Christmas Day came 
the first troops of the Americal Division, 
the 164th Regimental Combat Team. 
Bidden farewell by one of the area's fre- 
quent earthquakes, the battle-weary 3d 
Marines departed on the ships that had 
carried the 164th. On 28 December Gen- 
eral Hodge arrived and took over com- 
mand of the eastern sector from Turn- 
age, and the i82d Regimental Combat 
Team prepared to take over from the 
21st Marines. The i32d Regimental 
Combat Team took over its part of the 
line on 9 February, and five days later 
the Americal's field artillery battalions, 
the 221st, 245th, 246th, and 247th, began 
relieving the 3d Division's artillery regi- 
ment, the 12th Marines. 14 The 3d De- 
fense Battalion and several Marine air 
squadrons remained at Empress Augusta 

With the Japanese quiescent in De- 
cember except for intermittent air at- 
tacks at night, the immediate problems 
facing Griswold were logistical rather 
than tactical. The road net had to be 
finished; a good road net would not 

" Total casualties for the I Marine Amphibious 
Corps to 15 December were: 

1. Empress Augusta Bay, 293 killed, 1,071 wound- 
ed, 95 missing, and 1,161 sick and evacuated. (The 
relatively large figure for missing was due to the 
McKean's sinking and the loss of many of her 

2. Treasuries, 53 killed, 174 wounded, and 1 

3. Choiseul, 7 killed, 14 wounded, 4 missing. In 
addition 1 Marine Amphibious Corps lost several 
men at the base depot in Vella Lavella to aerial 
bombardment. Of the total casualties, the 3d Marine 
Division lost the most— 186 killed, 624 wounded. 
The 37th Division suffered 10 wounded during the 
period. All these figures are taken from a casualty 
report in I Marine Amphibious Corps' report, Phase 
III. The 3d Marine Division reported that it had 
counted a, 100 dead Japanese. 

4.2-Inch Chemical Mortar firing in 
support of infantry troops. 

only improve the supply situation but 
would give Griswold all the benefits of 
interior lines if and when the Japanese 
attacked in strength. The inland air- 
fields had to be completed, beach con- 
gestion ended, more dumps and depots 
established. General Griswold stated the 
problem thus: 

Puruata Island was so heavily loaded 
down it was about to sink. All beaches were 
congested. No long range supply road sys- 
tem had been planned. Long hand carry 
was the rule, particularly in the Marine Di- 
vision sector (later the Americal) for the 
front line troops. Forward ration dumps, 
ammunition and bomb dumps, gasoline 
dumps, hospital areas and bomb shelters for 
the same, beach developments, interior sup- 
ply roads, the Service Command area itself, 
a central cemetery, refrigeration, sawmills, 
drainage ditches, and a myriad of other 



Lt. Gen. Millard F. Harmon, center, and Maj. Gen. Oscar W. Griswold, second 
from [eft, are briefed by a Marine officer. 

ig44. By 31 January its strength was 
slightly more than two thousand men. 

Logistical development under Gris- 
wold was extensive and orderly. By now 
the swamp had been drained. Malaria 
was kept rigidly under control. The 
volcanic ash of the region made adequate 
roads, but the heavy rainstorms that fell 
almost daily tended to wash them away. 
Road maintenance was therefore one of 
the most difficult logistic problems. By 
1 March forty-three miles of two-way 
and thirty-six miles of one-way roads had 
been built. The troops also cleared 
several acres for gardens. The hot sun 
and frequent rains gave them fair re- 
turns, and fresh vegetables, normally a 
rarity in that part of the world, improved 
the otherwise almost unvarying diet of 
C and K rations and dehydrated foods. 

things were non-existent, and not even 
visualized. Space for all these things had to 
be carved out of the virgin jungle. 15 

Griswold, characterized by Halsey as 
"a farsighted and capable planner," set 
to work with his staff. 16 Harmon's head- 
quarters contributed greatly to the solu- 
tion of the logistical problems by activat- 
ing, in New Caledonia on 15 December, 
a Provisional Service Command for Bou- 
gainville. This organization, specifically 
tailored for the particular mission of 
supporting the XIV Corps, began its 
operations on Bougainville on 6 January 

1s Ltr, Griswold to Barnett, CofS USAFISPA, 15 
Feb 44, quoted in SOPACBACOM, The Bougain- 
ville Campaign, Ch. 5, p. 239, OCMH. 

18 Halsey, Narrative Account of the South Pacific 
Campaign, p. n, OCMH. 



Green vegetables grew fairly well, but 
tomatoes and corn did not. 17 There were 
frequent distributions of books, movies, 
performances by motion picture and 
radio personalities, sports, and occasion- 
ally beer and soft drinks. Empress 
Augusta Bay was about as pleasant a 
beachhead as one could hope for. 

During: the first two and a half months 
following Griswold's assumption of com- 
mand there was no heavy fighting. There 
were not enough troops to hold all the 
high ground inland, but combat and 
reconnaissance patrols went out to the 
east, the north, and the west to keep tab 
on all the possible routes of Japanese 
approach. Airplanes also reconnoitered 
trails, and PT boats, water routes. 

One of the outstanding patrols was 
conducted by the ist Battalion of the 
Fiji Infantry Regiment, which arrived in 
late December. This battalion, composed 
of 34 officers (some white, some Fijian) 
and *j*j>j enlisted Fijians, was at first com- 
manded by Lt. Col. J. B. K. Taylor of 
New Zealand. But Taylor was wounded 
his first night ashore and was replaced 
by Maj. Geoffrey T. Upton, also of New 
Zealand. 18 A detachment of the Fiji bat- 
talion left the beachhead on 28 Decem- 
ber and walked through the mountains 
over the Numa Numa Trail to the vil- 

" Ltr, Hon. Hugh M. Milton, II (former CofS, 
XIV Corps), to amhor, 13 Jul 56, OCMH. Excep- 
tions to the unvarying diet were: turkey for 
Thanksgiving and Christmas, steak early in 1944, 
and fresh (cold storage) eggs on Easter Sunday. 

18 The Fijians added more color to the beach- 
head than did any other unit. Immaculate in ap- 
pearance, they were nearly all men of extraordinary 
physical stature. They obviously liked soldiering 
and their marching was impressive in its precision. 
They sang well and often, their repertoire ranging 
from their native songs through "Onward Christian 
Soldiers" to "Pistol Packin' Mama" in Fijian. 

lage of Ibu, east of the mountains, where 
they set up an outpost on 30 December. 
From Ibu these natural jungle fighters 
kept watch over enemy movements on 
the east coast so that no Japanese could 
advance unsuspected along the Numa 
Numa Trail. They reported to corps 
headquarters by radio and were supplied 
by air drops from C-47's. They also 
hacked an airstrip suitable for L-4 planes 
(Piper Cubs) on the 1,700-foot-high shelf 
that Ibu rests on. 

But during December 1943 all ground 
operations were of minor importance 
when compared with the air operations 
against Rabaul that were conducted by 
South Pacific aircraft. 

December Attacks Against Rabaul 

Eight Seabee battalions and one New 
Zealand engineer brigade had begun 
work on a fighter strip at Cape Torokina 
promptly after D Day. Because the area 
was one of the few relatively dry patches 
of ground at Empress Augusta Bay, there 
was some competition among other units 
to occupy it, but the squatters were 
evicted and the builders were able to 
work unimpeded. 19 The strip was ready 
for operations on 9 December, and the 
next day seventeen F4U's of Marine 
Fighting Squadron 216 (I Marine Air- 
craft Wing) flew in and set up at their 
new base. 

Starting in mid-November B-24's of 
the Thirteenth Air Force had begun 
bombing Rabaul every few days, but 

" This section is based on Craven and Cate, The 
Pacific: Guadalcanal to Saipnn, pp. 350-52; Mori- 
son, Breaking the Bismarcks Barrier, pp. 392-98; 
USSBS, The Allied Campaign Against Rabaul, 
passim. See also Sherrod, History of Marine Corps 
Ax'iation in World War II, pp. 193-97. 



C-47 Air-Dropping Supplies on a partially completed airstrip. 

now, with its new forward fighter fields 
at Torokina and in the Treasuries, Air 
Command, Solomons, was ready to start 
an intensive series of operations with 
the purpose of completely neutralizing 

The Solomons air command now had 
a new commander. Maj. Gen. Ralph J. 
Mitchell of the Marine Corps relieved 
General Twining on 20 November. 
Twining returned to the United States, 
then went to Italy where he commanded 
the Fifteenth Air Force. 20 The strength 
of General Mitchell's command was 
formidable, even after the intensive 
operations of October and November. 
He had, in operating condition on 17 
December, igg fighters, 200 light and 
medium bombers, and 99 heavy bomb- 

30 In the closing days of the war, Twining led the 
Twentieth Ait Force. 

ers, or about twice what the Japanese 
had in the entire Southeast Area. 21 

The first time the Torokina field was 
used against Rabaul was 17 December, 
when fighters from New Georgia staged 
through it on a 76-plane sweep. From 
then on it was almost continuously in 
use. For the rest of December, except 
when the weather was too bad for flying, 
Mitchell continued the attacks, varying 
fighter sweeps with fighter-escorted raids 
by B-24's. 22 

But Mitchell's heavy bomber pilots, 

51 Total strength of Mitchell's command, includ- 
ing nonoperational planes, was 268 fighters, 252 
light and medium bombers, and 1 1 1 heavy bombers. 

11 On Christmas Day Halsey sent a carrier raid 
against Kavieng. He ordered a surface bombard- 
ment of Buka in order to lure out enemy aircraft, 
whereupon Admiral Sherman's two carriers (Bunker 
Hill and Monterey) struck Kavieng soon after sun- 
rise. But Sherman's pilots found few targets. 



like Kenney's, were unable to knock out 
Rabaul, and toward the end of the year 
the Japanese sent in more planes. Me- 
dium, dive, and torpedo bombers 
would have to be used, and their em- 
ployment would have to await comple- 
tion of the strips near the Piva River. 
The first of these, termed Piva U or Piva 
Uncle, was started on 29 November and 
completed on 30 December. The second, 
Piva Yoke, was ready on g January 
1944- 23 

11 These dates are taken from Building the Naxty's 
Bases in World War II, pp. 270-72. 

It was clear that the reduction of 
Rabaul would not occur until 1944. 
Kenney's and Mitchell's attacks in 1943, 
however, were quite effective, if not com- 
pletely successful. They caused enough 
damage to make the Japanese garrison 
start wholesale excavation in November 
in an effort to put everything possible 
under ground and so escape complete 

Meanwhile, under partial cover of the 
invasion of Bougainville and Mitchell's 
attacks on Rabaul, General MacArthur's 
forces had crossed Vitiaz and Dampier 
Straits to invade New Britain. 


Crossing the Straits 

Plans and Preparations 

By November 1943 Cartwheel was 
rolling along rapidly and smoothly. In 
just over five months Nassau Bay, Wood- 
lark, Kiriwina, New Georgia, Vella La- 
vella, Salamaua, Nadzab, Lae, the Mark- 
ham Valley, Finschhafen, the Treasuries, 
and Empress Augusta Bay had fallen to 
the Allies. At the newly won bases air- 
fields were either in operation or under 
construction. Allied planes dominated 
the skies all the way to Rabaul, and 
Allied ships sailed the Solomon Sea and 
the Huon Gulf in comparative safety. 

The capture of Finschhafen in Opera- 
tion II was a step toward control of the 
straits between New Guinea and New 
Britain, a control that would help make 
possible the drive toward the Vogelkop 
Peninsula and the Philippines in 1944 
and would be essential to any amphibi- 
ous advance against Rabaul. The South- 
west Pacific's next move (Operation III 
of Elkton III) was first planned by 
GHQ on the assumption that Rabaul 
would be captured. Looking eastward 
rather than westward from the Huon 
Peninsula, it aimed at the establishment 
of air forces at Cape Gloucester on 
western New Britain and of PT boat 
bases on the south coast of New Britain. 

Allied control over Rabaul and over 
Kavieng and Manus in the Admiralties, 
and to provide bases on the north side of 
the straits to insure the safe passage of 

Selecting Targets 

GHQ's orders for the operation, given 
the code name Dexterity, were pub- 
lished on 22 September. 1 They directed 
General Krueger's Alamo Force, for- 
merly the New Britain Force, supported 
by Allied Air and Allied Naval Forces, 
and by U.S. Army Services of Supply, 
Southwest Pacific, to seize Cape Glou- 
cester by airborne and amphibious in- 
vasions and to neutralize the forward 
Japanese base at Gasmata on southern 
New Britain, to gain control over 
western New Britain as far east as the 
line Gasmata-Talasea, and to capture 
Vitu and Long Islands beyond the straits. 
General Blarney's New Guinea Force 
would meanwhile continue its opera- 
tions in the Huon Peninsula and the 
valleys. Gasmata was to be neutralized 
by troops who would land at nearby 
Lindenhafen Plantation, establish an 
emergency airfield, and advance on Gas- 
mata in a shore-to-shore movement. The 

(See Map 12.) These were to increase 

1 GHQ SWPA OI 38, 28 Sep 43, and amendments, 
in GHQ SWPA G-3 Jnl, 22 Sep 43. 



plan directed Krueger to prepare to par- 
ticipate with South Pacific forces in the 
capture of Rabaul but this order was 
canceled on 10 November. 2 Saidor was 
not specifically mentioned, although 
both MacArthur and Chamberlin had 
suggested it as a target earlier in the 

D Day for the invasion of Cape 
Gloucester was initially set for 20 No- 
vember but was postponed twice. The 
final date was 26 December. 

This plan provoked a good deal of 
disagreement. The first to protest was 
General Kenney. With the decision to 
bypass Rabaul obviously in mind, he 
presented his objections to General Mac- 
Arthur on 10 October. The original con- 
cept, he argued, called for an encircling 
ring of air bases, including Cape Glou- 
cester, to be established around Rabaul 
in order to lay siege to it. But now that 
"faster action is contemplated" it would 
take too long to develop Gloucester into 
a useful air base. It would not be neces- 
sary to take either Gloucester or Linden- 
hafen, he told MacArthur: bases at Do- 
bodura, Nadzab, and Kiriwina, plus the 
one at Finschhafen and perhaps a new 
one at Saidor, could provide support 
for invasions as far away as Kavieng,* In 
speaking of faster action, Kenney appar- 
ently was referring to the long-range 
plan Reno III, which was then being 
prepared. It called for completion of the 

*GHQ SWPA OI 38/12, 10 Nov 43. Apparently 
there was a strategic lag at GHQ so that the impact 
of the Joint Chiefs' order to bypass Rabaul was not 
fully reflected at once in orders prepared by the 
GHQ staff. 

3 Ltr, Kenney to CINCSWPA, 10 Oct 43, sub: 
GHQ OI 38, aa Sep 43, in GHQ SWI'A G-3 Jnl, 11 
Oct 43; Kenney, General Kenney Reports, pp. 326- 
27; General Kenncy's comments on draft MS of 
this volume, OCMH. 

Cartwheel operations and then the 
move toward the Philippines, according 
to the following schedule: Hansa Bay, 1 
February 1944; Kavieng (by the South 
Pacific), 1 March 1944; Admiralties, 1 
March 1944; neutralization of Rabaul 
and perhaps, later, its occupation; Hum- 
boldt Bay and Arafura Sea, 1 June 1944; 
Geelvink Bay-Vogelkop Peninsula, 15 
August-i October, 1944; Halmahera, 
Amboina, the Palaus, 1 December 1944; 
Mindanao, 1 February 1945. 

General Chamberlin, MacArthur's 
G-3, observed that the air general's plan 
differed from MacArthur's present 
plans. 4 There would be time, he asserted, 
to complete airdromes at Gloucester be- 
fore undertaking the next operations. 
While he did not state that the Kavieng 
invasion could not be supported without 
Gloucester, he pointed out several ad- 
vantages to be derived from the move to 
New Britain: 

1. The Allies could better control 
Vitiaz Strait. (Here he reversed himself 
on the position he had taken on the 
point the month before.) Control from 
one side would be possible but it would 
be dangerous to leave the other side in 
Japanese hands, 

2. Cape Gloucester would provide 
better support for the Kavieng and Ad- 
miralties attacks provided for in Reno 

3. Cape Gloucester would provide 
better cover for convoys moving through 
Vitiaz Strait against the Admiralties. 
Even assuming the bypassing of Rabaul, 
Chamberlin concluded, a point on the 
south coast of New Britain would be 
needed to control Vitiaz Strait, neutralize 

4 Chamberlin must also have been referring to III. 



Gasmata, and provide an emergency air- 
field for planes attacking Rabaul. 5 

Admirals Carpender and Barbey also 
seem to have favored holding both sides 
of the straits, as did General Krueger. 
The admirals did not favor the seizure 
of Gasmata, because they felt it would 
mean a reckless exposure of ships to 
Rabaul-based aircraft. 6 

Kenney was informed that Mac- 
Arthur's plans, which encompassed the 
bypassing of Rabaul, required Cape 
Gloucester and Lindenhafen. 7 But Ken- 
ney's argument, coupled with the ad- 
mirals' and added to the facts that Gas- 
mata was swampy and that the Japanese 
were known to be sending more troops 
there, did have some effect. One month 
later MacArthur canceled Gasmata 
operations and directed the Alamo 
Force to seize Cape Gloucester and to 
establish control over adjacent islands 
and "minimum portions" of western 
New Britain with the purpose of pro- 
tecting Cape Gloucester. 8 

The matter did not end there. It was 
finally settled by a conference at GHQ 
in Brisbane on 21 November attended by 
Kenney, Carpender, and Barbey. The 
naval commanders opposed Gasmata and 
are reported to have wanted a PT boat 

Memo, ACofS G- 3 GHQ SWT A for CofS GHQ 
SWPA, 11 Oct 43, sub: Comments on Ltr From 
Comdr AAF, 10 Oct 43, in GHQ SWPA G-3 Jnl, 
11 Oct 43; Craven and Gate, The Pacific: Guadal- 
canal to Saipan, pp. 330-31. 

"Ltr, Adm Barbey to Gen Smith, Chief of Mil 
Hist, 20 Nov 53, no sub, OCMH. 

7 Ltr, GHQ SWPA to Comdr AAF, 16 Oct 43, 
sub: GHQ OI 38, in GHQ SWPA G- 3 Jnl, 16 Oct 

8 GHQ SWPA OI 38/12, 10 Nov 43. General 
Whitehead also disliked both Gasmata and Cape 
Gloucester. See his letter to Kenney, 1 1 November 
1943, quoted in Graven and Cate, The Pacific- 
Guadalcanal to Saipan, pp. 329-30. 

base elsewhere on New Britain's south 
coast. Therefore Arawe, the name of a 
peninsula, a harbor, and an island west 
of Gasmata which had been listed as an 
objective in Elkton III, was substituted 
for Gasmata with the intention of using 
it as a PT base and in the hope of divert- 
ing the enemy's attention from Cape 
Gloucester. Arawe had a fair anchorage 
and there were only a few Japanese in 
the area. General Kenney assured his 
fellow commanders that he could give 
better air cover to Arawe than to Gas- 
mata. Cape Gloucester remained the 
main objective. 9 As the same ships had 
to be used for both invasions, the dates 
were staggered. 

Setting Dates 

The first dates selected, 14 November 
for Lindenhafen and 20 November for 
Cape Gloucester, proved impossible to 
meet and had to be postponed. The 
process of postponement and selection of 
new dates clearly illustrates some of the 
controlling factors in Southwest Pacific 
amphibious operations. 

By 26 October Kenney, Sutherland, 
and Chamberlin realized that enough 
air cover would not be available to meet 
the first target dates. The Finschhafen 
airstrip would not be completed until 
about 5 December. Construction of the 
Lae-Nadzab road had fallen behind 

"Memo, SJC [Chamberlin] for MacArthur, 21 
Nov 43, sub: Practicality of Establishing PT Base 
S Coast Western New Britain, in GHQ SWPA G-3 
Jnl, 21 Nov 43; GHQ OI 38/15, 22 Nov 43; Ltr, 
Gen Krueger to Gen Smith, Chief of Mil Hist, 31 
Oct 53, no sub, OCMH. For conflicting accounts 
see Kenney, General Kenney Reports, pp. 326—27, 
and Morison, Breaking the liismurcks Barrier, p. 
372. After the war Barbey expressed the view that 
the PT base was never an important factor. 



schedule and it could not take heavy 
vehicles and machinery before 1 Decem- 
ber; consequently the three airstrips in 
the lower Markham Valley would not be 
in shape to maintain air operations be- 
fore 15 December. The VII Amphibious 
Force, which would carry the assault 
troops in Dexterity, could not be re- 
leased from its responsibilities for sup- 
plying Lae and Finschhafen for some 
time. It was estimated that from 135,000 
to 150,000 more tons of supplies would 
have to be sent to Lae, 60,000 to 70,000 
more to Finschhafen, in order to sup- 
port air operations. Shipments to Nad- 
zab were slowed by the lack of enough 
men and docks at Lae, and movement 
of supplies to Finschhafen was slowed 
by the fact that until the airfield was 
finished the naval commanders would 
not risk sending heavy ships there. 

Southwest Pacific invasions usually 
took place during the dark of the moon 
to help hide ships from nocturnal raid- 
ing planes. The last-quarter moon would 
come on ig November, the first-quarter 
moon on 4 December. If the attack could 
not be mounted before 4 December it 
would have to be put off until after ig 
December, the date of the next last- 
quarter moon. But this was the period 
of the northwest monsoon, and the 
longer the Southwest Pacific waited for 
ideal moon conditions the rougher 
would be the surf at Cape Gloucester. 

Chamberlin therefore recommended 
that Dexterity be put off until the 
earliest possible date in December, that 
the VII Amphibious Force keep on sup- 
plying Lae and Finschhafen a while 
longer, and that two engineer aviation 
battalions that were scheduled for Cape 
Gloucester be set to work at Lae and 

Finschhafen for the time being. 10 Mac- 
Arthur, accepting these recommenda- 
tions, announced that he would delay 
the attack about fifteen days, and that 
the VII Amphibious Force would sup- 
ply Lae and Finschhafen until about 20 
November. 11 

This decision provoked the quiet, un- 
dramatic General Krueger to protest that 
the resulting schedule would be too 
tight. MacArthur's order meant that 
Gloucester would have to be invaded on 
4 December. The subsidiary operation 
would have to be accomplished on 28 
November. Since there was no reserve 
shipping, any losses on 28 November 
would hamper the main landing. 
Further, the VII Amphibious Force, 
once relieved at Lae and Finschhafen, 
could not be expected to get to Milne 
Bay until 26 or 27 November. Thus 
there would hardly be time for re- 
hearsals. Krueger, asking for more ships 
or for more time, suggested that the first 
operation take place on 2 December, 
Gloucester on the 26th. 12 MacArthur 
agreed to another postponement and 
eventually set Z Day for Arawe at 15 
December, D Day for Gloucester at 26 
December. 13 

ALAMO Force Plans 

Originally assigned to Alamo Force 
for Dexterity were the 1st Marine Di- 
vision; the 32d Division; the 6g2d Tank 

10 Memo, Chamberlin for CofS GHQ SWPA, 26 
Oct 43, sub: Data of Attack on New Britain, in 
GHQ SWPA G-3 Jnl, 26 Oct 43. 

" Rad, MacArthur to Comdrs NGF, Alamo, et 
al, 28 Oct 43, in GHQ SWPA G-3 Jnl, 28 Oct 43. 

12 Ltr, Krueger to MacArthur, 12 Nov 43, no sub, 
in GHQ SWPA G-3 Jnl, 14 Nov 43. 

13 GHQ SWPA OI 38/17, 3 Dec 43; GHQ SWPA 
OI 38/18, 16 Dec 43, 



Destroyer Battalion; the 503d Parachute 
Infantry Regiment, for Cape Gloucester 
only; and a number of quartermaster, 
medical, signal, engineer, and antiair- 
craft units. The 1st Cavalry and 24th 
Infantry Divisions, then in Australia but 
soon to move to New Guinea, and the 
503d Parachute Infantry (which would 
be committed at Cape Gloucester) con- 
stituted GHQ's reserve. The 1st Marine 
and 3 2d Divisions moved from Australia 
to the forward area shortly before the 
invasions. 11 

As usual, MacArthur gave Krueger 
responsibility for co-ordinating the plans 
of supporting air and naval forces with 
those of the Alamo Force. In contrast 
with the system of unity of command 
over all elements of an invasion force 
that prevailed in the South Pacific, the 
commander in chief specifically directed 
that Allied Air and Naval Forces would 
operate under GHQ through their re- 
spective commanders and exempted 
them from control by Alamo or New 
Guinea Forces. However, if the Japanese 
attacked in any area the senior local com- 
mander was to control all Southwest 
Pacific forces in the threatened area. 

General Krueger and the Alamo 
Force staff had been planning for Dex- 
terity since August. 15 In the beginning 
Alamo headquarters was at Milne Bay, 
where it had been established at the 
opening of the Cartwheel offensives. 

14 The 32d Division went to Milne Bay and Good- 
enough, the 1st Division to Milne Bay, Oro Bay, 
and Goodenough. 

15 Elkton III had provided for the invasion of 
Cape Gloucester, Arawe, and Gasmata, and in 
August and September GHQ had prepared general 
plans and specific orders that were superseded by 
Reno III and OI 38. See GHQ's Marfa Plans in 
Alamo Force G-3 Jnl Dexterity No. 1. 

On 21 October Krueger moved it to 
Goodenough Island. 16 

During the planning period for Dex- 
terity the Japanese were strengthening 
their garrisons in western New Britain 
in accordance with the orders issued by 
Imamura in September. Thus Allied 
estimates of Japanese strength in the 
area rose from 500 before September to 
2,500 on the 26th. In December Krueger 
placed enemy strength at between 5,668 
and 9,344, with the strongest concentra- 
tion at Cape Gloucester. The 1st Marine 
Division, apparently deriving its infor- 
mation from the same sources as Alamo 
Force, arrived at a higher figure— be- 
tween 8,400 and 12,076." 

Little was known about the terrain of 
western New Britain, and Krueger 
ordered ground reconnaissance in addi- 
tion to the extensive air photography 
that w^s undertaken by Allied Air 
Forces. Because PT boats were not al- 
lowed to operate off New Britain's north 
coast no patrols were able to examine 
Borgen Bay, where the main Cape 
Gloucester landings were to take place. 
Marine patrols landed from PT boats 
and reconnoitered the area south of Cape 
Gloucester from 24 September through 

16 Sixth Army headquarters remained at Camp 
Columbia near Brisbane until 2 February ig44 
when it moved to Cape Cretin on the southeast 
corner of the Huon Peninsula. Alamo headquarters 
had moved from Goodenough to the cape on 24 
December to be near the scene of operations. The 
advance echelon of GHQ remained at Port Moresby. 

11 Alamo Force Rpt, Dexterity Opn, 15 Dec 43— 
10 Feb 44, and Incls, 17 May 44; Alamo G-2 
Periodic Rpt 18, 9 Dec 43, in Alamo Force G-3 Jnl 
Dexterity No. 6; 1st Mar Div Order of Battle and 
Strength Est, New Britain, 13 Dec 43, in Alamo 
Force G-3 Jnl Dexterity No. 7. The order of battle 
lists were nearly correct except that the Allies did 
not know that the iyth Division was moving from 
Rabaul to western New Britain. 



21 December in a series of patrols. A 
group of Alamo Scouts, an informal re- 
connaissance organization operating di- 
rectly under General Krueger, recon- 
noitered Gasmata from 6 through 27 
October. On the night of g-10 December 
one American officer and five natives 
disembarked from a PT boat east of 
Arawe, scouted the area, and concluded 
there were only a few Japanese present. 18 
More information was obtained from 
aerial photography. Missions were flown 
almost daily so that Alamo and sub- 
ordinate headquarters could be kept in- 
formed of gun positions, beach defenses, 
bridges, and trails. The VII Amphibious 
Force used air photos as the basis for its 
hydrographic charts, and the 1st Marine 
Division used them to pick the landing 
beaches. 10 

Krueger's first tactical plans, prepared 
in accordance with GHQ's orders, had 
called for the heavily reinforced 126th 
Regimental Combat Team, under Brig. 
Gen. Clarence A. Martin, of the 32d Di- 
vision, to take Gasmata. Cape Gloucester 
was to have been captured by the Back- 
hander Task Force under Maj. Gen. 
William H. Rupertus, commander of the 
1st Marine Division. The assault force 
was to have consisted of one regimental 
combat and one battalion landing team 
of Rupertus* division, the 503d Para- 
chute Infantry Regiment, and the 12th 
Marine Defense Battalion. The marines 
were to have delivered an amphibious 

18 The patrols on western New Britain included 
Maj. John V. Mather. Australian Army; Sub-Lt. 
Andrew Kirkwcll-Smith, a coastwatcher in the Aus- 
tralian Navy; and Sub-Lt. William G. Wiederaan, 
also of the Australian Navy but before the war a 
Church of England missionary at Sag Sag on 
western New Britain. 

19 Alamo Force Rpt, Dexterity Opn; Craven and 
Cate, The Pacific: Guadalcanal to Saipan, p. 332. 

assault, coupled with a parachute jump 
by the 503d. 20 But this whole plan was 
drastically revised. 

When on 22 November General Mac- 
Arthur substituted Arawe for Gasmata, 
Krueger decided to use a smaller force 
than the 126th. He correctly believed 
Arawe to be weakly defended. For Arawe 
he formed the Director Task Force 
under Brig. Gen. Julian W. Cunning- 
ham, who as a colonel had led the in- 
vasion of Woodlark. Its assault units in- 
cluded Col. Alexander M. Miller's two- 
squadron 112th Cavalry; the 148th Field 
Artillery Battalion; the 59th Engineer 
Company; Headquarters and Headquar- 
ters Battery, 236th Antiaircraft Artillery 
(Searchlight) Battalion; and C and D 
Batteries, 470th Antiaircraft Artillery 
(Automatic Weapons) Battalion. In re- 
serve was the 2d Battalion, 158th In- 
fantry. Supporting garrison units, to be 
moved to Arawe after 15 December (Z 
Day), consisted of several engineer, medi- 
cal, ordnance, and other service detach- 
ments. All these units had been attached 
to the Alamo Force for the invasion of 
the Trobriands in June, and were still 
occupying the islands. 

The concept of the Cape Gloucester 
invasion was changed also; the parachute 
jump was canceled and the 503d re- 
moved from the troop list. Several fac- 
tors contributed to this change. General 
Krueger's headquarters had never liked 
the idea. General Rupertus, too, had op- 

In reserve was Maj. Gen. William H. Gill's g2d 
Division less the 126th Regimental Combat Team. 
General Chamberlin douhted the wisdom of using 
the 503d as no suitable drop zone was to be found. 
Alamo Force Rpt, Dexterity Opn; Alamo Plan of 
Opn, Dexterity, 28 Sep 43, in GHQ SWPA G-g 
Jnt, 28 Sep 43; Memo, ACofS C-g CHQ SWPA for 
CofS GHQ SWPA, 10 Oct 43, sub: Plan of Opns 
Dexterity Submitted by Alamo Force, same file. 



posed the parachute jump from the start. 
He pointed out that bad weather, which 
had prevented several air attacks against 
Rabaul, might interfere with the para- 
chute jump and thus deprive him of a 
substantial part of his assault force. 21 
General Kenney's headquarters, in De- 
cember, added its opposition. First, al- 
though Alamo Force orders did not 
specify exactly how the jump was to be 
accomplished, it was understood at 
Allied Air Forces headquarters that a 
piecemeal and therefore dangerous drop 
was planned. Second, it seemed that the 
jump would be under way about the 
time that Japanese planes might be ex- 
pected to turn up. Asking if the jump 
was necessary, Kenney's operations officer 
stated emphatically that the air com- 
mander wanted "no part" of it. 22 

General Rupertus' headquarters had 
disliked the whole scheme of maneuver 
as prescribed by Alamo headquarters, 
as well as the parachute jump. Alamo's 
first plans called for simultaneous, sepa- 
rated landings by two small forces, which 
were to converge on the airfield at Cape 
Gloucester in conjunction with the 
5ogd's jump. But the ist Marine Di- 
vision, which had had ample experience 
with jungle warfare on Guadalcanal, 
felt that this plan was unsound because 
the rough and scarcely known terrain 
could easily delay either or both of the 
marching forces. Also, the Japanese 
could be expected to outnumber any one 
of the three landing forces. When Gen- 
erals MacArthur and Krueger visited ist 

21 Memo, CG Backhander TF for CG Alamo, 3 
Nov 43, no sub, in Alamo Force G-3 Jnl Dexterity 
No. 3. 

22 Check Sheet, Dir Opns AAF for G_3 GHQ 
SWPA, 8 Dec 43, in GHQ SWPA G-3 Jnl, 8 Dec 43. 

Marine Division headquarters at Good- 
enough on 14 December, Col. Edwin A. 
Pollock, divisional operations officer, 
frankly expressed the marines' objections 
to the parachute jump and the scheme of 

Krueger had included the parachute 
jump because MacArthur's headquarters 
had assigned the 503d Parachute In- 
fantry to the operation, and he con- 
sidered himself under orders to make 
his plans fit the forces assigned. 23 Mac- 
Arthur, Krueger, and Kenney now dis- 
cussed the matter further. It developed 
that Dobodura would not support the 
mounting of the 503d as well as all the 
planned bomber operations. To use the 
503d would require moving one heavy 
bomber group from Dobodura to Port 
Moresby, and bad weather over the Owen 
Stanleys might keep the bomber group 
out of action. The jump was canceled. 

Alamo Force further revised its tac- 
tical plans for taking Cape Gloucester 
to meet the ist Marine Division's ob- 
jections. Final plans called for one regi- 
mental combat team to land on two 
beaches on the north coast of New 
Britain between Silimati Point in Bor- 
gen Bay and the airfield at Cape Glouces- 
ter, while a second (less a battalion 
landing team) landed immediately be- 
hind, passed through the first, and at- 
tacked toward Cape Gloucester to the 
airfields. One battalion landing team was 
to land near Tauali on the west coast of 
New Britain to block the coastal trail 

23 Lt. Col. Frank O. Hough, USMCR, and Maj. 
John A. Crown, USMCR, The Campaign on New 
Britain (Washington, 1952), p. 19; Gen Krueger's 
comments on the draft MS of this volume, attached 
to his ltr to Maj Gen A. C. Smith, Chief Mil Hist, 
31 Oct 53, no sub, OCMH. 



and prevent reinforcement of the air- 
drome area from the south or retreat of 
the airdrome garrison to the south. 

The assault units of Rupertus' Back- 
hander Task Force were two regimental 
combat teams of the ist Marine Di- 
vision; the 12th Marine Defense Battal- 
ion less its 155-mm. gun group; detach- 
ments, including LCM's and LCVP's, 
of the 2d Engineer Special Brigade; and 
the 913th Engineer Aviation Battalion. 
The reserve, supporting, and garrison 
units included the remainder of the 1st 
Marine Division, the 155-mm. gun group 
of the iath Defense Battalion, and a large 
number of engineer, medical, quarter- 
master, and malaria control units, chiefly 
of the Army. 

The Arawe (Director) forces were 
to mount the invasion at Goodenough, 
the Gloucester (Backhander) forces 
through Oro Bay, Goodenough, and 
Milne Bay. In Alamo reserve was the 
32d Division. 24 

Logistical plans called for the U.S. 
Army Services of Supply, Southwest 
Pacific Area, now commanded by Maj. 
Gen. James L. Frink, to establish and 
maintain at New Guinea bases sixty 
days' supply of all types except chemical 
and Air Forced 5 Thirty days' of the last 
two classes were to be maintained. 
Frink's command was to make building 
materials for ports and air bases available 

"* See Alamo Force, Rpt, Dexterity Opn; Alamo 
FO 5, and annexes, 30 Nov 43, in Alamo Force G— 3 
Jnl Dexterity No. 5; Amendment 1 to Alamo 
FO 5, 15 Dec 43, in Alamo Force G— 3 Jnl Dex- 
terity No. 7; Craven and Cate, The Pacific: Guadal- 
canal to Saipan, p. 331; Hough and Crown, The 
Campaign on New Britain, p. jg. See also ACofS 
G-3 Alamo, Revised G-3 Study, Gloucester, 2 Dec 
43, in Alamo Force G— g Jnl Dexterity No. 5. 

M General Marshall was Deputy Chief of Staff, 

to the Alamo Force by Z plus 5 and D 
plus 5, and was to furnish naval forces 
with supplies common to the Army and 
Navy pending establishment of the naval 
supply system, or in emergencies. The 
VII Amphibious Force would of course 
transport supplies to the beachheads un- 
til they were secured and Frink was 
ready to take over. Allied Air Forces 
was to transport supplies to the ground 
troops in emergencies. 

All units in the task forces were to be 
stripped of equipment not needed for 
their combat missions. They would carry 
to Arawe and Cape Gloucester in the 
assault echelons thirty days' supply and 
three units of fire, ivhich would be built 
up by succeeding shipments to sixty 
days' supply and six units of fire (ten 
for antiaircraft). Oro Bay was the main 
supply base, Milne Bay the secondary. 28 
Cape Cretin, near Finschhafen on the 
southeast corner of the Huon Peninsula, 
which the Alamo Force was preparing 
as a supply point and staging base, was 
to serve for resupply. 

Krueger, on receiving data on Allied 
Air Forces' requirements, directed the 
Backhander Task Force to build a small 
strip at Cape Gloucester for air supply 
at once, a ioo-by-5, 000-foot runway by 
D plus 10; a second ioo-by-5,ooo-foot 
runway, capable of expansion to 6,000 
feet, by D plus 30; and also overruns, 
parallel taxi ways, roads, and airdrome fa- 
cilities. 27 

x GHQ SWPA OI 38, 22 Sep 43, in GHQ SWPA 
G-3 Jnl, 22 Sep 43; Alamo Administrative Order 4, 
30 Nov 43, in Alamo Force G-3 Jnl Dexterity 
No. 5. 

" Appendix 4, Annex 4, Engr, to Alamo FO 5, 
17 Dec 43, in Alamo Force G— 3 Jnl Dexterity 
No. 8. 



The Enemy 

When the Allies landed on Cape 
Gloucester General Imamura could not 
have been surprised. He had anticipated 
such a move some time before. In Oc- 
tober the 8 th Area Army staff had con- 
cluded that two lines of action were 
open to the Allies: capture of the New 
Britain side of the straits, invasion of 
Bougainville, and a direct assault upon 
Rabaul in February or March 1944; or 
the slower process of isolating Rabaul 
by seizing the Admiralties and Kavieng. 28 
Considering the first course the more 
likely, he decided to send more troops 
to western New Britain in addition to 
those he had sent under General Mat- 
suda in September. He would have 
liked to reinforce the Admiralties and 
Kavieng but felt he could not spare any 
more troops from the defenses of Rabaul. 

Imamura therefore ordered the ijth 
Division, less the battalions dispatched 
to Bougainville, to western New Britain. 
Reaching Rabaul from China between 4 
October and 12 November, the 17 th 
went by echelons to its new posts by 
naval vessel and small boat. The move- 
ment began in October but was still 
under way in mid-December. 

The zyth Division commander vvas 
given operational control of the units 
already there, chiefly General Matsuda's 
65th Brigade and 4th Shipping Group 
at Cape Gloucester and 2d Battalion, 
228th Infantry, and two naval guard 

!8 This subsection is based on 8th Area Army 
Operations, Japanese Monogr No. 1 10 (OCMH), 
pp. 85— 1 19; Southeast Area Air Operations, 1942— 
44, Japanese Monogr No. 38 (OCMH), pp. 25-29; 
Southeast Area Naval Operations, III, Japanese 
Monogr No. 50 (OCMH), pp. 36—39; 17th Division 
Operations in Western New Britain, Japanese 
Monogr No. 1 1 1 (OCMH), pp. 1-7. 

companies at Gasmata. 29 Final plans or- 
ganized the entire force into three com- 
mands. The first and largest, under Mat- 
suda, consisted of the 65th Brigade 
(principally the 141st Infantry), the 4th 
Shipping Group, and a large number of 
field artillery, antiaircraft, automatic 
weapons, engineer, and communications 
units. Matsuda was charged with defense 
of the area from the emergency airstrip 
and barge staging point near Tuluvu 
around the coast to Cape Busching on 
the south. Under Matsuda Maj. Masa- 
mitsu Komori, with most of the 1st Bat- 
talion, 81st Infantry, one company of 
the 34th Infantry, and engineers plus 
detachments, was assigned responsibility 
for defense of Arawe. Col. Shuhei Hira- 
shima, commanding the 34th Infantry, 
less the 2d Battalion, the 2d Battalion, 
228th Infantry, the 2d Battalion, 23d 
Field Artillery Regiment, and the naval 
guard companies, was to hold the air- 
field at Gasmata. 30 The ijth Division 
established its command post at Gavuvu, 
east of the Willaumez Peninsula and a 
long distance from the scene of opera- 

The air strength available to the Jap- 
anese for the forthcoming fight had 
been reduced, not only by the Allied air 
attacks against Rabaul, but also by orders 
from Tokyo. The 2d Area Army was 
established in the Netherlands Indies on 
1 December 1943, and the boundary be- 
tween it and Imamura's 8th Area Army 
was set at longitude 140 east. But the 

29 The 65th Brigade had played an important 
part in the fighting on Luzon in the first Philip- 
pines campaign. See Morton, The Fall of the 

30 A false report from natives in October that the 
Allies had landed east of Gasmata had caused the 
dispatch of the naval companies to Gasmata. 

B-25's Over Wewak. Three damaged enemy planes are visible on the ground. 



yth Air Division was transferred out of 
the 4th Air Army and assigned to the 
2d Area Army. 31 This transfer seriously 
reduced Imamura's forces, but so far the 
yth Air Division had not operated effec- 
tively. Its most outstanding exploit had 
been the loss of planes on the ground 
at Wewak. 

Air Operations 

With the new fields in the Markham 
Valley and at Finschhafen in operation, 
Allied Air Forces' aerial preparations f6r 
Dexterity were the most extensive yet 
seen in the Southwest Pacific. 32 They in- 
cluded, besides daily P-38 photographic 
missions, long-range search missions by 
PBY's of the Seventh Fleet's Patrol 
Wing io, RAAF Catalinas, and Fifth 
Air Force B-24's, and bombing and 

Air attacks, which had been under 
way against New Britain since October, 
began on a large scale in late November. 
Cape Gloucester and Gasmata were the 
main targets. Arawe was avoided until 
14 December in order to keep from 
warning the Japanese. During December 
Kenney's planes attacked Gasmata or 
Gloucester, or both, nearly every day 
and sometimes twice a day. As General 
Whitehead said, Cape Gloucester was 
"tailor made" for air operations. The 
target area lay along the beach and was 
long and narrow. 33 During December 
Kenney's planes flew 1,845 sort ies over 

Sl Smith, The Approach to the Philippines, Ch. 

1,2 This subsection is based on Craven and Cate, 
The Pacific: Guadalcanal to Saipan, pp. 333-38. 

33 Memo, Lt Col Paul Weyrauch, Asst Arty Off 
Alamo, and Lt Col Carl A. Fields, Air Off Alamo, 
for CofS Alamo, 10 Nov 43, no sub, in Alamo Force 
G— 3 Jnl Dexterity No. 3. 

Gloucester, dropped 3,926 tons of bombs, 
and fired 2,095,488 rounds of machine 
gun ammunition. The chief targets were 
Tuluvu airfield, antiaircraft guns, sup- 
ply dumps, and the barge staging points. 
The airfield was knocked out of action 
early in the operation and stayed that 



The 11 2th Cavalry, shipped aboard 
LST's, reached Goodenough Island from 
Woodlark on 1 and 2 December. 34 There 
General Cunningham gave Colonel 
Miller detailed orders for the landing 
of his regiment at Arawe on 15 Decem- 

Arawe, which before the war had been 
a regular port of call for vessels of the 

M This section is based on Ltr, Comdr VII Amphib 
Force to COMINCH, 10 Jan 44, sub: Arawe Opn, 
and incls, in GHQ SWPA G-3 Jnl, 10 Jan 44; 
Alamo Force Rpt, Dexterity Opn, and Incl i. 
Lessons Learned; Director TF, Hist Rpt, Arawe 
[in form of ltr, Gen Cunningham to CG Sixth 
Army, 6 June 44]; 112th Cav, Hist Rpt [Arawe], 
24 Nov 43-10 Feb 44; 8th Area Army Operations, 
Japanese Monogr No. 110 (OCMH), pp. 119-22; 
17th Division Operations in Western New Britain, 
Japanese Monogr No. 111 (OCMH), pp. 8-14; 
Southeast Area Air Operations, Japanese Monogr 
No. 38 (OCMH), pp. 29-30; Memo, Capt Joseph 
H. Baker for CO 5g2d Engr Boat and Shore Regt, 
18 Dec 43, no sub, in Alamo Force G-3 Jnl Dex- 
terity No. 9; Alamo Fragmentary FO's 1 and 2, 
27 Dec 43, in Alamo Force G-3 Jnl Dexterity No. 
1 1 ; Capt T. H. Baker, USMC, Rpt Amphib Tractor 
Opn Arawe, 27 Dec 43, in Alamo Force G-3 Jnl 
Dexterity No. 12; Ltr, Gen Cunningham to CG 
Alamo, 6 Jan 44, sub: Opns Director TF, in 
Alamo Force G-3 Jnl Dexterity No. 15; Ltr, CG 
2d ESB to CTF 76, 16 Dec 43, sub: Rpt Arawe 
Landing, and Log of Events as Seen From SC 742, 
both in Alamo Force G-3 Jnl Dexterity No. 8; 
CTF 76 Opn Plan A3-43, 10 Dec 43, and Memo, 
JWC [Cunningham] for Gen Krueger, 18 Dec 43, 
no sub, and 2d ESB FO 1, 11 Dec 43, all in Alamo 
Force G-3 Jnl Dexterity No. 7; G-2 Alamo Ter- 

Burns-Philp South Seas Company, had 
a harbor suitable for large vessels. There 
were several beaches that landing craft 
could use, of which the two best were 
House Fireman on the west coast of the 
boot-shaped Arawe peninsula and the 
village of Umtingalu on the mainland, 

rain Rpt, Arawe, g6 Nov 43, and Ltr, Hq Alamo 
to CC Backhander TF, 30 Nov 43, no sub, and 
Director TF FO t, 4 Dec 43, and Memo, DAA 
[Maj D. A. Alberti, Alamo G-3 Sec] for G-3 
Alamo, 8 Dec 43, sub: Observation Director Exer- 
cise, all in Alamo Force G— 3 Jnl Dexterity No. 5; 
Off of Chief Engr, GHQ AFPAC, Critique, pp. 
109-17; Craven and Cate, The Pacific: Guadal- 
canal to Saipan, p. 335; Hough and Crown, The 
Campaign on New Britain, pp. 140-52; Isely and 
Crowl, The U.S. Marines and Amphibious War, pp. 
18G-88; Morison, Breaking the Bismarcks Barrier, 
PP- 373-77- 

seventeen hundred yards east of the 
peninsula's base. The rest of the coast 
line consisted of stone cliffs about two 
hundred feet high, interspersed with low 
ground that was covered by mangrove 
swamp. Reefs fringed all the beaches, 
and it was clear that LCVP's could not 
get to the shore until detailed reconnais- 
sance for passages was made. (Map ij) 
Therefore General Krueger arranged 
with Rupertus for one company of the 
1st Marine Amphibian Tractor Battal- 
ion to be attached to the Director Task 
Force to take the assault waves ashore. 
Krueger also attached part of the 592d 
Engineer Boat and Shore Regiment, 2d 
Engineer Special Brigade, with 17 
LCVP's, 9 LCM's, 2 rocket-firing 



DUKW's, and 1 repair and salvage boat, 
to Task Force 76 for the landings. 35 

The 112th Cavalry stayed on Good- 
enough for ten days. During this period 
the troops received additional practice 
and training with all their weapons, in- 
cluding two new ones— the flame thrower 
and the 2.36-inch rocket launcher (ba- 
zooka). Before shoving off all men were 
informed of the general plan of attack 
and given aerial photographs and maps 
to study. The training period was topped 
by two landings. The first was intended 
to familiarize the troops with loading 
and unloading landing craft. The second 
was conducted under assumed combat 
conditions and involved the co-ordinated 
landing of all elements at proper inter- 
vals and their tactical deployment 
ashore. General Cunningham forcefully 
pointed out several major deficiencies. 
Units were not always under control of 
their commanders, intervals between 
landing waves were too long, and not all 
junior officers and noncommissioned of- 
ficers knew their duties. 

With Generals MacArthur and Krue- 
ger looking on, the Director Task 
Force boarded the LSD Carter Hall, 
HMAS Westralia, and the APD's Sands 
and Humphreys on the afternoon of 13 
December. 30 At midnight the ships de- 

00 The rocket DUK.W, with 4.5-inch rockets, was 
an experimental craft that the engineer special 
brigades of the Southwest Pacific Area had de- 
veloped in an effort to provide fire support for 
landings after naval gunfire had ceased or lifted. 
These DUKW's attempted, without much success, 
to carry out the function performed by the LCI 
gunboat in the South Pacific. 

™ The Westralia is listed variously as an A PA 
and an LSI. TF 76, for Arawe, including support- 
ing echelons, consisted of the ships listed above, 
ten escort and bombardment destroyers, an escort 
and mine group of patrol craft and subchasers, 

parted for Buna, where General Cun- 
ningham left the Carter Hall and joined 
Admiral Barbey aboard the destroyer 
Conyngham, the flagship. The voyage to 
the target, which included a feint to- 
ward Finschhafen, was uneventful eK- 
cept for seas rough enough to cause 
the passenger troops some discomfort. 
Admiral Crutchley's cruisers and de- 
stroyers covered the move to the east 
while PT boats patrolled the straits to 
the westward. 

The Landings 

Barbey's convoy sighted the south 
coast of New Britain shortly after 0300, 
15 December, and the troop ships soon 
hove to in the transport area about five 
miles east of Arawe. 37 By 0450 the Carter 
Hall had launched thirty-nine loaded 
amphibian tractors bearing the assault 
waves and the two rocket DUKW's out 
of her well deck. 

Dawn was still one hour away when 
150 men of A Troop, 112th, who had 
been aboard the APD Sands, started for 
the beach at Umtingalu in fifteen rub- 
ber boats. They had been ordered to 
make a surprise landing in darkness at 
H minus 1 hour and block the coastal 
trail that was the Japanese escape and 
reinforcement route to the east. About 
0525, when the boats were nearing shore 
and in the moonlight were probably vis- 
ible from the shore, they came under 
fire from machine guns, rifles, and a 25- 
mm. dual purpose gun, which prompt- 
ly sank all but three of the rubber boats. 

several LCT flotillas, and a service group of LST's 
plus landing craft attached from the ist Marine 
Division and the 2d Engineer Special Brigade. 

31 All times are approximate. All sources employed 
give differing times for the same events. 



The fire continued while the troops 
floundered in the water divesting them- 
selves of their light combat packs and 
outer clothing. The destroyer Shaw then 
opened fire and quickly silenced the 
enemy. 38 Small boats picked up the sur- 
vivors of A Troop, who later landed 
without arms and almost naked at 
House Fireman Beach. Twelve men 
w T ere killed, four missing, and seventeen 
wounded in this repulse. 

B Troop fared better. Ordered to land 
at H minus 1 hour on the islet of 
Pilelo, across Pilelo passage from the 
peninsula, its men were to take the 
Japanese by surprise and silence a radio 
station that was reported at the village 
of Paligmete. They left the APD Hum- 
phreys on fifteen rubber boats at the 
same time that A Troop left the Sands. 

B Troop had planned to surprise the 
enemy by landing at Paligmete village, 
but when the Japanese started firing 
on A Troop it was obvious that surprise 
was lost. B Troop landed at Wabmete, 
on the west of Pilelo, instead. Once 
ashore the cavalrymen started on foot 
for Winguru. 

The leading platoon reached Win- 
guru at 0615 and met fire from Japanese 
in two caves on the rising ground south 
of the village. Leaving one squad to con- 
tain these Japanese, B Troop pushed 
on to Paligmete, found neither Japanese 
nor radio, and returned to Winguru to 

38 General Cunningham was wroth at the Shaw's 
delay in opening fire, but Admiral Barbey and 
Brig. Gen. William F. Heavey [Commanding Gen- 
eral, 2d Engineer Special Brigade], who observed 
the operation from the deck of SC 742, reported 
that the Shaui held her fire because she could not 
immediately locate any targets. The boats and the 
shore, viewed from the sea, blended into a dark 
blur. General Cunningham had tried the predawn 
landing against Admiral Barbey's advice. 

mop up. Bazooka fire closed one cave 
but the other was faced with logs which 
proved impervious to rockets and ma- 
chine guns. Finally a flame thrower 
team, covered by machine gun fire, 
edged to within fifteen yards of the cave 
and let loose a blast of flame. B Troop 
then moved in, tossed grenades, and 
the action was over. One American sol- 
dier had been killed. Seven dead enemy 
were found. The action ended about 

Meanwhile the main landing at 
House Fireman Beach had been accom- 
plished successfully if not flawlessly. The 
assault waves came from Lt. Col. Clyde 
E. Grant's 2d Squadron, 112th Cavalry, 
organized into five landing waves: ten 
LVT (A) (s)'s (Buffaloes), carrying E 
and F Troops, in the first; eight LVT 
(i)'s (Alligators) each in the second, 
third, and fourth waves; and five Alli- 
gators in the fifth. The waves were 
scheduled to land at five-minute inter- 
vals. H Hour was set for 0630, after 
the conclusion of the air and naval bom- 
bardments. One and a half hours were 
allowed for the amphibian tractors to 
proceed from the ships to the beach, a 
move which would take place in poor 
light. Since dawn came at 0624 and sun- 
rise at 0646, the landing itself would 
take place in daylight. 

But someone along the line had be- 
come confused. Once boated, the first 
wave started directly for the shore in the 
dark. Brig. Gen. William F. Heavey, 
commanding the 2d Engineer Special 
Brigade, who had come along as an ob- 
server aboard the landing wave control 
craft, SC 742, saw the boats dimly about 
0500. When radio communication with 
the flagship unaccountably failed, the 



subchaser's captain and Heavey headed 
off the errant amphibian tractors. There 
was much confusion and milling about 
in the darkness, and it was 0600 or later 
before the tractors regained their forma- 

Destroyers bombarded House Fire- 
man Beach with 1,800 5-inch rounds 
from 0S10 to 0625, whereupon B-25's 
took over. Three squadrons had been 
assigned to air alert over Arawe under 
control of an air liaison party aboard 
the Conyngham, and the first of these 
bombed and strafed the peninsula and 
the beach. Under ideal conditions the 
interval between the cessation or lifting 
of support bombardment and the land- 
ing of troops is only long enough to 
prevent the troops from being hit by 
their own support fire, but the lead wave 
of tractors had been slowed by the con- 
fusion and by a stiff current in Pilelo 
passage. It did not land until after 0700. 
On the way in, the wave met machine 
gun fire that was quickly silenced by 
4.5-inch rockets from the control craft 
and the two rocket DUKW's on the 
flanks. Otherwise there was no opposi- 

This was fortunate, because the suc- 
ceeding waves in the Alligators, which 
were slower craft than the Buffaloes, had 
not been able to keep up. Twenty-five 
minutes elapsed before the second wave 
landed, ample time for a resolute de- 
fending force to have inflicted heavy 
casualties on the first wave. When an- 
other fifteen minutes had passed the last 
three waves came ashore practically to- 
gether. 39 

39 After the event aU units and observers report- 
ing on the subject declared it a mistake to use 
vehicles with differing speeds in the assault waves. 

The ad Squadron, once landed, re- 
organized, sent patrols to the toe of 
the peninsula, and pushed northwest- 
ward toward the base against slight op- 
position from scattered riflemen and 
rear guards. E Troop located twenty or 
more Japanese in caves in the cliff on 
the east side of the peninsula, killed 
several, and passed on. When others 
came out of their caves to snipe and 
harass, the 11 2th Cavalry Headquarters 
Troop sent out a patrol which disposed 
of them. 

Only two companies of Japanese sol- 
diers had been in the area, and when 
the 2d Squadron came ashore they re- 
treated eastward. Major Komori and his 
force had not yet reached Arawe. 

Meanwhile passages through the reefs 
had been found. The reserve 1st Squad- 
ron, under Maj. Harry E. Werner, had 
debarked from the Westralia while the 
Carter Hall was launching the tractors. 
Werner's squadron came ashore about 
0800 in the 2d Engineer Special Bri- 
gade's 2 LCM's and 17 LCVP's. An hour 
later Barbey's second echelon, 5 LCT's 
carrying 150 tons of gear and 50 men 
per LCT, and 7 LCM's carrying 25 tons 
of gear per LCM, arrived from Cape 
Cretin and began unloading. 

Operations at the beach were not 
smooth. The detachments forming the 
shore party had never worked together 
before, and although the beach was a 
good one it soon became congested. 
There was room for but two LCT's at 
one time; so unloading of beaching craft 
continued all day. 

For Dexterity the admirals had won 
the air cover argument, and planes were 
assigned as combat air patrol over the 



ships instead of standing by on ground 
alert. The first fighter cover, in the form 
of 8 P-38's, took station overhead at 
0715. This cover was subsequently in- 
creased and was maintained all day but 
it was not able to prevent an air attack 
at ogoo. The nth Air Fleet at Rabaul 
had just received more planes and now 
totaled 50 bombers and 100 fighters. 
Both Kusaka's fleet and the 6th Air Divi- 
sion sent out planes against Arawe. One 
flight of these, reported as consisting 
of 20 or go planes, eluded the P-38's 
and delivered the attack at 0900. The 
Westralia and Carter Hall, unloaded be- 
fore dawn, had departed at 0500 to 
avoid air attack, The rest of Task Force 
76, with the exception of craft actually 
at the beach and the flagship Conyng- 
ham, which remained to direct opera- 
tions, sought the cover of clouds and 
rain squalls. The Japanese bombed and 
strafed the beached LCT's, the Conyng- 
ham, and the troops for about five min- 
utes, scored no hits, and left with P-38's 
in pursuit. 

By midafternoon the Director Task 
Force controlled the entire peninsula. 
The 2d Squadron had reached the base, 
and now began establishing a main line 
of resistance there. Over sixteen hun- 
dred men, five hundred from the at- 
tached units and the rest from the two 
squadrons of the 112th Cavalry, were 

Operations, 16 December 194}- 
10 February 1944 

During the next few days LCT's from 
Cape Cretin and APD's from Good- 
enough brought in heavy weapons, sup- 
plies, and more troops. There was no 

ground contact with the Japanese at 
Arawe, but in the air the enemy reacted 
with violence. Between 15 and 27 De- 
cember naval planes delivered seven at- 
tacks against Arawe and against the 1st 
Marine Division at Cape Gloucester, 
and in about the same period the 6th 
Air Division attacked four times. LCT's 
at Arawe on 16 December suffered al- 
most continuous air attack. Resupply 
convoys lost one coastal transport sunk 
and another damaged, plus one mine- 
sweeper and six LST's damaged. Al- 
though General Cunningham's force had 
no go-mm. antiaircraft guns to keep 
bombing planes away, damage ashore 
was fortunately light. Cunningham ex- 
pressed his urgent need for the go-mm. 's, 
but none was available for Arawe. By 
late December, however, the nth Air 
Fleet and the 6th Air Division had lost 
so many planes to Allied fighters over 
New Britain, to Southwest Pacific at- 
tacks against Wewak, and to the South 
Pacific's raids on Rabaul that they were 
forced to stop daylight bombardment 
and confine their activities to the defense 
of Rabaul and Wewak. When Imamura 
asked Tokyo for more planes Imperial 
Headquarters responded by sending the 
8th Air Brigade to Hollandia under the 
2d Area Army. 

The Japanese had not yet given up 
at Arawe, When the ijth Division com- 
mander received word of Cunningham's 
landing he ordered Major Komori, who 
was then proceeding by boat and over- 
land march to Arawe from Rabaul, to 
make haste. He also ordered the 1st 
Battalion, 141st Infantry, to move from 
Cape Busching to Arawe and come un- 
der Komori's command. Komori was 



then to destroy the Director Task 

The Americans soon became aware of 
the approaching Japanese. On 18 De- 
cember two Japanese armed barges at- 
tacked a 112th Cavalry patrol on board 
two of the 2d Engineer Special Brigade's 
LCVP's (which had remained under 
Cunningham at Arawe). The Japanese 
scored hits; the patrol abandoned the 
LCVP's and made its way east to Arawe 
on foot. Komori's force reached the 
Pulie River east of Arawe on 20 Decem- 
ber, advanced west, and on Christinas 
Day forced the 112th to abandon its 
observation posts and outposts east of 

Cunningham, correctly concluding 
that the Japanese were converging 
against him from two directions but er- 
roneously concluding that Komori's 
command was but the advance guard of 
a stronger force from Gasmata, asked 
Krueger for reinforcements. The Alamo 
commander hastily dispatched G Com- 
pany, 158th Infantry, by PT boat. 

Komori, with the 1st Battalion, 81st 
Infantry, reached the area northwest of 
the main line of resistance on 26 De- 
cember. Like the Americans, he had dif- 
ficulty in getting any exact informa- 
tion on positions in the featureless, jun- 
gled terrain at the peninsula's base. Sev- 
eral of his night probing attacks were 
repulsed by mortar fire, as were daytime 
attacks on 28 and 29 December. The sec- 
ond of these took the lives of most of 
his men, but the 1st Battalion, 141st 
Infantry, arrived in the late afternoon 
of 29 December. Several small attacks in 
early January 1944 by the 112th were 
beaten off, but the cavalrymen estab- 

lished the fact that the Japanese were 
digging in about six hundred yards be- 
yond their own perimeter. Komori had 
resolved to defend the prewar airstrip 
on the mainland east of the peninsula, 
which in any case the Allies did not 

On 6 January Cunningham reported 
to Krueger the existence of the Japanese 
positions. Cunningham's forces now to- 
taled almost 4,750 men and his short 
front line— seven hundred yards— was a 
strong position with fields of fire cut, 
barbed wire emplaced, and artillery and 
mortar data computed. 41 The enemy po- 
sitions he faced consisted largely of shal- 
low trenches and foxholes and were 
practically invisible in the dense under- 
brush. There were only about 100 Japa- 
nese and half-a-dozen machine guns 
there, but lack of visibility and the fact 
that the Japanese moved their guns fre- 
quently made them almost impossible 
for artillery and mortars to hit. An as- 
sault would be further complicated by 
the fact that in the area there were no 
clearly defined terrain features which 
could serve to guide an attack and help it 
maintain its direction. Cunningham 
asked for tanks and more troops, and 
repeated his request for 90-mm. antiair- 
craft guns. 

Krueger agreed that attacks by rifle- 
men alone against Komori would result 
in a waste of lives and agreed to send 

10 For an amusing incident about this airstrip and 
the embarrassment caused by overenthusiasm on 
the part of public information officers, see Hough 
and Crown, The Campaign on New Britain, pp. 

41 By 10 January Task Force 76 had carried 4,750 
men and 8,165 tons of supplies to Arawe. 



tanks as well as more troops. F Com- 
pany, 158th Infantry, and B Company, 
1st Tank. Battalion, 1st Marine Division, 
reached Arawe from Cape Cretin on 10 
and 12 January. 

On the morning of 16 January attack 
and medium bombers struck at the Jap- 
anese positions, artillery and mortars 
shelled them, and the Marine light 
tanks, two companies of the 158th Infan- 
try, and C Troop of the 112th Cavalry 
attacked. The tanks led, with infantry- 
men and cavalrymen following each 
tank. Direct communication between 
tanks and foot troops was successfully 
attained by a device which the tank 
company improvised; it installed an 
EE8 field telephone at the rear of each 
machine. The attack went well and car- 
ried forward for fifteen hundred yards. 
Next day B Troop and one tank pla- 
toon mopped up remaining pockets of 

Thereafter Arawe was quiet. Casual- 
ties for all units in the Director Task 
Force totaled 118 dead, 352 wounded, 
and 4 missing. Komori had actually 
withdrawn to defend the airstrip, and 
remained there until ordered to retreat 
to the east in February. 

Ironically enough, no PT base was 
ever built at Arawe. Actually the final 
plans had never included any provision 
for one. Lt. Comdr. Morton C. Mumma, 
commanding Southwest Pacific PT's, 
successfully insisted that he did not want 
and did not need a PT base there to 
patrol the straits or to attack Japanese 
barges, which seldom used the south 
coast anyway. Arawe never became an 
air base either. The only airstrip ever 
used was one for artillery liaison planes 

Alligator Returning to Beach on 
Arawe for more supplies, 18 Decem- 
ber 1943- 

that engineers hastily cleared on 13 Jan- 

Cape Gloucester 

Meanwhile the main event at Cape 
Gloucester had gotten under way. Ele- 
ments of the 1st Marine Division sched- 
uled for the main assault landings east 
of Cape Gloucester conducted final re- 
hearsals at Cape Sudest on 21 Decern- 



ber. 42 The heavily reinforced 7th Ma- 
rines boarded ship at Oro Bay three days 
later and departed at 0600 on Christmas 
morning. En route ships carrying the re- 
inforced 1st Marines (less one battalion 
landing team) from Cape Cretin joined 
up. The convoy then made its way peace- 
fully through Vitiaz Strait, sailed be- 
tween Rooke and Sakar Islands, and ap- 
proached Cape Gloucester. The 2d Bat- 
talion Landing Team, 1st Marines, em- 
barked at Cape Cretin and steamed 
through Dampier Strait for Tauali. 13 Ad- 
miral Crutchley's Task Force 74— the 
American cruisers Phoenix and Nash- 
ville, HMAS Australia and HMAS 
Shropshire, and eight destroyers— es- 
corted Task Force 76 while PT boats 
patrolled the northern and western en- 
trances to the straits. 

The Landings 

In the dim light of 0600 on 26 De- 
cember Crutchley's ships opened their 
supporting bombardment on the landing 
beaches east of Cape Gloucester, a bom- 
bardment that continued for ninety 

45 This section is based on Craven and Cate, The 
Pacific: Guadalcanal to Saipan, pp. 337—45; Hough 
and Crown, The Campaign on New Britain; Mori- 
son, Breaking the Bismarcks Barrier, pp. 378-89; 
Office of the Chief Engineer, GHQ AFPAC, En- 
gineers of the Southwest Pacific: 1941-1945, VI, 
Airfield and Base Development (Washington, 1951), 
192-95; Alamo Force Rpt, Dexterity Opn; 8th 
Area Army Operations, Japanese Monogr No. 110 
(OCMH), pp. 119-22; 17th Division Operations in 
Western New Britain, Japanese Monogr No. 111 
(OCMH), pp. 14-21, 

43 For Cape Gloucester Task Force 76 consisted of 
the flagship Conyngham, 10 APD's, 16 LCI's, 12 
destroyers, g minesweepers, and 24 LST's; 14 LCM's, 
12 LCT's, and 2 rocket DUKW's went to Tauali. 
In reserve were the Westralia and Carter Hall. De- 
tachments of the 2d Engineer Special Brigade— 181 
men, 33 landing craft, and 2 rocket DUKW's— were 
attached to the 1st Marine Division. 

minutes. Two new LCI's equipped with 
4.5-inch rockets took station on the flanks 
as guide and fire support craft. After 
threading their way through a difficult 
channel, APD's, in the lead, lowered 
landing craft full of troops while behind 
them LCI's and LST's awaited their 

turns at the beaches. (Map iS) 

The 1st Air Task Force of the Fifth 
Air Force had prepared extensive plans 
for all-day air cover and support that 
involved a total of five fighter squadrons 
and fourteen attack, medium, and heavy 
bomber squadrons. The first support 
bombers arrived from Dobodura about 
0700 and B-24's, B-25's, and A-20's 
bombed and strafed the beaches and the 
airdrome. B-25's dropped smoke bombs 
on Target Hill, the 450-foot ridge just 
west of Silimati Point that gave clear ob- 
servation of the beaches and airfields. 
A-20's strafed the landing beaches until 
the leading wave of landing craft was 
five hundred yards from shore. At that 
time the naval gunfire was moved inland 
and to the flanks. 

An errant breeze blew so much smoke 
from Target Hill that some of the lead- 
ing waves of landing craft carrying the 
7th Marines could not easily identify the 
beaches. There was no opposition at the 
proper beaches, where most of two bat- 
talions of the 7th landed, but a detach- 
ment which wandered three hundred 
yards too far west- had a brisk fire fight 
on shore. 

The 7th Marines found that the land- 
ing beaches were good but very shallow. 
And as the assault waves crossed the 
beaches they were brought up sharply 
by jungle so dense they had to start hack- 
ing to get inland. Immediately behind 
the beach was a narrow shelf of relatively 



dry ground. Behind the shelf was a 
swamp which made anything like rapid 
movement or maneuver completely im- 
possible. Men floundered through the 
mud, slipping into sinkholes up to their 
waists and even their armpits. And in the 
swamp giant trees, rotted by water and 
weakened by bombs and shells, toppled 
over easily. The first marine fatality on 
Cape Gloucester was caused by a falling 

The narrow beach and the swampy 
jungle behind it caused a good deal of 
congestion, especially when the LST's 
began discharging their cargo. As ex- 
pected, Japanese planes from Rabaul at- 
tacked the ships and the beach during 
the day, although their first attacks were 

directed against Arawe in the belief that 
the convoy had been intended as rein- 
forcement for Cunningham. They sank 
one destroyer, seriously damaged two 
more, and scored hits on two others, as 
well as on two LST's. 

By the day's end the 7th Marines held 
the beachhead area. The artillery battal- 
ions of the 1 ith Marines had landed and 
emplaced their howitzers. The ist Ma- 
rines had come ashore, passed through 
the 7th, and begun the advance west 
toward the airdrome. The regiment first 
attempted to advance with battalions 
echeloned to the left rear, but the swamp 
forced movement in a long column with 
a narrow front along the coastal trail. 

Also on 26 December the reinforced 




26-29 December 1943 

Marine p£himeter,27 0ec 
Can/our interi/ol IOOO feel 

MAP 18 

2d Battalion, ist Marines, landed suc- 
cessfully at Tauali and D Company, 5g2d 
Engineer Boat and Shore Regiment, 2d 
Engineer Special Brigade, landed on 
Long Island to prepare a radar station 

The first night ashore at Gloucester 
was miserable, and it was the first of 
many more that were just as bad. 
Drenching rains characteristic of the 
northwest monsoon poured down in tor- 
rents; more trees fell. The Japanese in 
the airdrome area, estimating that only 
2,500 men had come ashore, counter- 
attacked the 7th Marines, but they 
failed, as did a heterogeneous group that 
later struck at the Tauali positions. 

Capture of the Airfield 

The 1st Marines started westward 
along the coastal trail toward the cape 
and the airfield at 0730, 27 December. 
The swamp still forced the regiment to 
advance in column of battalions, with 
the rear battalion echeloned as much as 
possible to the left rear. Each battalion 
marched in column of companies and 
sent small patrols into the swamp to pro- 
tect its flanks. General Sherman tanks of 
A Company, 1st Tank Battalion, sup- 
ported. By 1615, when it dug in for the 
night, the regiment had gained three 
miles, and had become aware of the 
existence of a large Japanese block about 
a thousand yards east of the airfield. 



7TH Marines Landing on Narrow Beach, Cape Gloucester, 26 December 1943. 

Next day the 1st made deliberate 
preparations before attacking the block. 
The absence of Japanese resistance the 
day before had led to the conclusion that 
the enemy was concentrating his forces 
inside the block. The 1st Marines waited 
during the morning for more tanks to 
make their way up the trail, which by 
now was a veritable morass, and for artil- 
lery and aircraft to shell and bomb the 
block. The infantry and the tanks moved 
to the attack about noon and shortly ran 
into the block. This position was a strong 
point which originally had faced the sea 
for defense of the beach but which served 
alternately as a trail block. It consisted 
of camouflaged bunkers with many anti- 
tank and 75-mm. weapons. There was 
scarcely room for tanks and infantry to 
maneuver, but by the end of the after- 

noon the 1st Marines had reduced the 

General Rupertus had been asking for 
his reserve, the reinforced 5th Marines, 
which had remained under Krueger's 
control, and on the 28th Krueger re- 
leased the regiment. Rupertus then de- 
cided to hold up the advance on the air- 
field until the 5th arrived. It came on 29 
December, but confusion over orders 
caused part of the 5th to land just behind 
the 1st, the rest at the D-Day beaches. 
When the 5th had been reassembled the 
drive began again. The 1st Marines con- 
tinued the coastal advance and, because 
the swampland on the left had given way 
to jungle, the 5th was able to make a 
wide southwesterly sweep. There was al- 
most no resistance. By the day's end most 
of the airdrome was in Allied hands and 



the major objective of the campaign had 
been achieved. 

The Japanese Withdraival 

This phase of the operation had gone 
rapidly and at the cost of comparatively 
few casualties. But the absence of Japa- 
nese opposition made it clear that a large 
body of the enemy must be elsewhere in 
the vicinity. Thus in the first two weeks 
of 1944 the 7th Marines and the 3d Bat- 
talion, 5th Marines, under Brig. Gen. 
Lemuel C. Shepherd, the assistant di- 
vision commander, attacked southward 
to clear Borgen Bay. Here waged the bit- 
terest fighting of the campaign as the 
141st Infantry struggled to keep posses- 
sion of the high ground. 

Thereafter there was little combat. 
The marines patroled extensively in 
search of the enemy, who proved to be 
elusive. B Company, 1st Marines, landed 
on Rooke Island on 12 February but 
found it had been evacuated. Eventually 
elements of Rupertus' division advanced 
by shore-to-shore movements as far east 
as Talasea without encountering any 
large numbers of Japanese. 44 

The enemy garrison at Cape Glouces- 
ter, especially Matsuda's command, had 
been in poor physical condition be- 
fore the invasion. The incessant air at- 
tacks against barge supply routes had 
forced it onto short rations, and malaria, 
dysentery, and fungus infections were 

The ijth Division commander, Lt. 
Gen. Yasushi Sakai, seems to have had 
little heart for a resolute defense, for he 
urged General Imamura to approve a re- 

44 In April the 40th Division relieved the 1st 
Marine Division and the 112th Cavalry on New 

treat to the Willaumez Peninsula. Ima- 
mura at first refused, then assented in 
January. On 23 February he ordered 
Sakai's entire command— ijth Division, 
6yth Brigade, 4th Shipping Group, and 
all attached units— to retreat all the way 
back to Rabaul and help defend it. 

Base Development 

Repair of the wrecked airdrome began 
in early January with the arrival of the 
913th Engineer Aviation Battalion. The 
work was complicated and slowed by 
jungle, rain, and swampy ground, and by 
nocturnal air raids that prevented night 
work for nearly two weeks. GHQ and the 
Fifth Air Force revised their require- 
ments and directed construction of a 
5,000-foot strip and a second, parallel 
strip 6,000 feet long. The 864th and 
141st Engineer Aviation Battalions ar- 
rived later in January and turned to on 
the strips and on roads and airdrome in- 
stallations. By the end of the month 4,200 
feet of the first airstrip had been covered 
with pierced steel matting, but it was 18 
March before the strip was completed. 
By then Rabaul, Kavieng, and the Ad- 
miralties had been neutralized or cap- 
tured and GHQ was planning the first 
giant step of its advance to the Philip- 
pines, a step which took it far beyond 
range of fighter planes based at Cape 
Gloucester. 45 The parallel 6,000-foot 
strip proved impossible to build and was 
never completed. 

Cape Gloucester never became an im- 
portant air base. It is clear that the 
Arawe and Cape Gloucester invasions 
were of less strategic importance than the 
other Cartwheel operations, and in the 

40 See Smith, The Approach to the Philippines, 
Ch. II. 



light of hindsight were probably not es- 
sential to the reduction of Rabaul or the 
approach to the Philippines. Yet they 
were neither completely fruitless nor ex- 
cessively high in casualties. The ist 
Marine Division scored a striking tactical 
success at the cost of 310 killed, 1,083 
wounded. And the Allied forces of the 
Southwest Pacific Area had, by means of 
these operations, broken out through the 
narrow straits. 


The first two Dexterity operations 
faced toward Rabaul, and as events later 
showed had much less effect on the 
course of the war than the other Cart- 
wheel operations. But in December 
1943 General MacArthur reversed his 
field and decided to exploit the tactical 
successes at Arawe and Cape Gloucester 
by moving west to seize Saidor, on the 
north coast of the Huon Peninsula. (See 
Map 12.) General Chamberlin had sug- 
gested Saidor in September, but it was 
1 1 December before an outline plan was 
prepared. 46 

" This section is based on Craven and Cate, The 
Pacific: Guadalcanal to Saipan, pp. 345-49; Mori- 
son, Breaking the Bismarcks Barrier, pp. 389-91; 
Off of Chief Engr, GHQ AFPAC, Engineers in 
Theater Operations, pp. 118—19, an( ' Critique, pp. 
118-32; Smith, The Approach to the Philippines, 
pp. 90-93; Alamo Force Rpt, Dexterity Opn; 
Alamo Force G-3 Jnl, Michaelmas Opn, Dex- 
terity; Michaelmas TF [126th RCT, reinfojeed], 
Rpt of Michaelmas Opn, 16 Dec 43-10 Feb 44; 
Michaelmas TF Opn Diary; CTF 76 Opn Plan 
4-43, 29 Dec 43; CTF Rpt of Saidor Opn, 3 Feb 44; 
GHQ SWPA Outline Plan Saidor, 10 Dec 43, in 
Alamo Force G-3 Jnl Dexterity No. 8; Memo, 
CAW [Gen Willoughby] for CofS GHQ SWPA, 21 
Dec 43, no sub, in GHQ SWPA G-3 Jnl, 21 Dec 43; 
GHQ Ltr of Instns to Comdrs Alamo Force, ANF, 
AAF, NGF, and USASOS, 17 Dec 43, sub; Michael- 
mas Opn, and GHQ SWPA OI 38/19, 31 Dec 43, 

Saidor lay slightly northeast of Mounts 
Gladstone and Disraeli, which glower at 
each other from their ii,ooo-foot emi- 
nences. It had a prewar airstrip, and had 
been used as a barge staging point by the 
Japanese. Lying no nautical miles from 
Finschhafen, 52 from Madang, and 414 
from Rabaul, it was well situated to sup- 
port the advance westward toward the 
Vogelkop and the move northward 
against the Admiralties. In addition 
Allied seizure would cut the 18th Army 
in two, for that army's main concentra- 
tions were at Madang-Wewak to the 
west, and at Sio-Gali to the east. 

Preparations and Plans 

The invasion of Saidor was not ac- 
tually decided on until 17 December, 
two days after the invasion of Arawe and 
nine days before the invasion of Cape 
Gloucester. On that date General Mac- 
Arthur ordered Krueger to prepare plans 
at once for an operation from Good- 
enough Island to seize Saidor and con- 
struct an advanced air and naval base 
there. Allied Air and Allied Naval Forces 
would support, and again MacArthur 
made Krueger responsible for co-ordina- 
tion of planning by the ground forces 
and the commanders of close support air 
and naval forces. The New Guinea 
Force, whose troops were now advancing 
against Sio and patrolling in the Ramu 
Valley beyond Dumpu, would support by 
continuing the move against Sio and by 
vigorous demonstrations in the Ramu 

both in GHQ SWPA G-3 Jnl, 22 Dec 43; ALF, Rpt 
on New Guinea Opns: 4 Sep 43—26 Apr 44; 8th 
Area Army Operations, Japanese Monogr No. 1 10 
(OCMH), pp. 86-91; l8tn Army Operations, II, 
Japanese Monogr No. 42 (OCMH), 125-80. 



Valley. U.S. Army Services of Supply 
was to haul supplies for the operation 
forward to Cape Cretin, where Krueger 
was to establish a temporary staging area 
and supply point pending the time that 
the U.S. Army Services of Supply base at 
Finschhafen began operating. Ground 
combat forces would come from the 
Alamo reserve for Cape Gloucester, the 
U.S. gad Division which had fought in 
the Papuan campaign. In addition Mac- 
Arthur assigned two engineer aviation 
battalions, an amphibian truck company, 
and an engineer boat and shore regi- 

Assignment of the mission to Alamo 
Force instead of New Guinea Force 
represented a departure from the prin- 
ciple that New Guinea Force would com- 
mand all operations in New Guinea. The 
change was probably made because 
nearly all trained Australian divisions 
were either committed to action or with- 
drawn for rest, and because it seemed 
clear that all the Alamo reserve would 
not have to be committed to Arawe or 

MacArthur gave Krueger and his air 
and naval colleagues little time to °;et 
ready. Actual initiation of the Saidor of- 
fensive, he announced, would depend on 
the progress of operations on Cape 
Gloucester, because landing craft for the 
former would have to come from the 
latter. It was expected, however, that D 
Day would be 2 January 1944, or shortly 
thereafter. Two days' notice would be 

GHQ's first outline plan had envisaged 
a parachute assault to take Saidor, but 
that was decided against because there 
still were not enough forward airfields 
to support current bombing operations 

and a parachute assault at the same time. 
The attack would have to be made by 
ground troops. 

For the attack Krueger organized the 
Michaelmas [Saidor] Task Force under 
command of General Martin, assistant 
commander of the g2d Division, who had 
just reached Goodenough Island. Most 
combat troops of the task force came out 
of the force Krueger had originally or- 
ganized for Gasmata. The rest were those 
assigned by General MacArthur. Mar- 
tin's force was built around the 126th 
Regimental Combat Team of the 32d 
Division, which included the 120th Field 
Artillery Battalion (105-mm howitzers). 47 
At the time of assignment the s;2d Di- 
vision units of the task force had just 
moved to Goodenough Island from 
Milne Bay. The rest of the force was 
scattered at such diverse points as Milne 
Bay, Kiriwina, and Lae. 

Although time for planning and pre- 
paring was short, and the pressure of 
Admiral Barbey's duties prevented him 
from conferring frequently with Krueger 
in person, the reports of the participating 
units bear witness to the fact that the 
experience and state of training of the 
commanders and troops were so high 
that things went smoothly. 

General Martin hastily organized a 
headquarters for his task force, taking as 
the nucleus Headquarters and Head- 
quarters Company, 126th Infantry. To 

"Other units were the 121st Field Artillery Bat- 
talion (75-mm. pack howitzers); Headquarters and 
Headquarters Battery, 191st Field Artillery Group; 
B and D Batteries, 209th Coast Artillery Battalion 
(AA, AW); A and D Batteries, 743d Coast Artillery 
Gun Battalion (AA); the Shore Battalion of the 
543d Engineer Boat and Shore Regiment, 2d En- 
gineer Special Brigade; the 808th and 863d En- 
gineer Aviation Battalions; and a variety o£ service 



Rear Adm. Daniel E. Barbey, left, and Brig. Gen. Clarence A. Martin, center, 
observing landing operations at Saidor, 2 January 1944. 

this were added some officers from g2d 
Division headquarters. Col. Charles D. 
Blanchard, task force G-g, came from 
Alamo Force as did a complete en- 
gineer section. Col. J. Sladen Bradley, 
commander of the 126th, also served as 
deputy commander and chief of staff of 
the Michaelmas Task Force. 

Plans began to take shape on 20 De- 
cember at a conference in Alamo head- 
quarters at Goodenough Island. Present 
were Admiral Barbey; Maj. Gen. 
William H. Gill of the 32d Division; 
General Heavey of the 2d Engineer Spe- 
cial Brigade; Brig. Gen. Edwin D. Pat- 
rick, Alamo Force chief of staff; Col. 
Clyde D. Eddleman, G-g of Alamo 

Force; General Whitehead; General 
Martin; and Colonel Blanchard. White- 
head, Patrick, and Eddleman presented 
their ideas, as did Barbey, who was just 
about to leave for Cape Gloucester. Both 
Barbey and Martin felt that, as Saidor 
was known to be lightly held by the 
enemy, preliminary air bombardment on 
the beaches on D Day would be of less 
value than a surprise landing at an early 
hour in the day. This conference was 
followed by two other brief ones during 
the next ten days. 

As there was neither time nor oppor- 
tunity for ground reconnaissance of 
Saidor, the landing beaches were chosen 
from aerial photographs. Three beaches, 



designated Red, White, and Blue from 
left (south) to right on the west shore of 
Dekays Bay just east of Saidor were se- 
lected. They were rough and stony but 
were chosen because they were close to 
the objective, because the beach gradient 
was steep enough to enable the troops to 
make a dry landing, because there was 
solid ground behind them, and because 
Dekays Bay could be expected to offer 
protection from the northwest monsoons 
prevailing at that time of year. 

Formal orders were published in late 
December. 48 Admiral Barbey organized 
his force generally as follows: 


Red Beach 3 

White Beach 2 

Blue Beach 4 


Red Beach 6 

White Beach 6 

Blue Beach 4 


Escort 4 

Bombardment 6 

Cover 5 

Control and rocket vessels: 

2 LCI's and 2 SC's. 
Supply group: 

6 LST's. 

These ships would carry the assault 
troops from Goodenough and land them 
at Saidor on D Day. Six more LST's 
would land additional troops and equip- 
ment on D plus i, and LST shipments 
would continue to bring in troops and 
supplies from Goodenough and Cape 
Cretin for some time thereafter. 

The assault waves of troops would land 

4S See Alamo [Escalator] FO 7, 22 Dec 43, and 
Michaelmas TF FO 1, 29 Dec 43, both in Alamo 
Force G— 3 Jnl File, Michaelmas; CTF 76 Opn Plan 
4-43, 29 Dec 43. 

from the APD's in thirty-six LCP(R)'s as 
follows: The 3d Battalion, 126th In- 
fantry, was to land on Red Beach at H 
Hour with two companies abreast, while 
the 2d Battalion, 126th, put one com- 
pany on White Beach and one company 
on Blue; the ist Battalion, 126th, would 
land from LCI's on White Beach at H 
plus 30 minutes. All units would push 
inland and reconnoiter. Field and anti- 
aircraft artillery were to land soon after 
the assault infantry. Forming the shore 
party would be A Company, 114th En- 
gineer Battalion; the Antitank Company 
and part of the Cannon Company, 126th 
Infantry; and the Shore Battalion, 542d 
Engineer Boat and Shore Regiment. In 
all, seven thousand men and three thou- 
sand tons of gear were to be put ashore 
on D Day. 

Plans for naval gunfire called for two 
destroyers to put 575 5-inch rounds of 
deep supporting fire at inland targets be- 
tween H minus 20 minutes and H Hour. 
Four destroyers would fire 1,150 rounds 
at the beach from H minus 20 to H 
minus 3 minutes when the lead wave of 
landing craft would be nine hundred 
yards from shore. Fire from rocket- 
equipped LCI's would cover the landing 
craft during the last nine hundred yards. 
In accordance with Martin's and Bar- 
bey's desire for surprise, the air plans did 
not provide for a preliminary bombard- 
ment on D Day. Provisions were made, 
however, for bombers to strike at inland 
targets after H Hour, and for strafers and 
fighters to execute supporting missions 
on call. 

H Hour was set for 0650, fifteen min- 
utes before sunrise— the earliest possible 
minute that would allow adequate light 
for the earlier naval bombardment. 



Weather could be expected to be either 
squally with rough surf or pleasant with 
smooth surf, and subject to sudden 
change from one extreme to the other. 

Troops of General Martin's task force, 
aside from the 32d Division units, were 
arriving at Goodenough while the plans 
were being prepared. Some units arrived 
short of clothing and equipment, and 
these were supplied as well as possible. 

Meanwhile General Krueger, con- 
cerned over the difficulties of supplying 
Arawe, Cape Gloucester, and Saidor 
simultaneously, argued in favor of post- 
poning Saidor. But General MacArthur, 
Vice Adm. Thomas C. Kinkaid, and 
others, promising to make sure that 
enough supplies arrived at Saidor, and 
unwilling to lose momentum, agreed 
that the operation would be valueless if 
postponed. Preparations went forward.^ 

Admiral Kinkaid, a cool, soft-spoken, 
bushy-eyebrowed product of many years 
in the U.S. Navy, had relieved Admiral 
Carpender as commander of Allied 
Naval Forces and of the Seventh Fleet in 
November. A classmate of Admiral 
Turner at the Naval Academy, he had 
already had ample experience in the 
Pacific, having commanded carrier task 
forces during the Guadalcanal Campaign 
and led the invasion of the Aleutians. 

On 28 December General Patrick 
notified Martin that the Cape Gloucester 
operation was proceeding successfully, 
and that his task force would probably 
invade Saidor on 2 January, the esti- 
mated date. As the LST's would have to 
sail from Goodenough on 31 December 

49 Ltr, Krueger to CINCSWPA, 28 Dec 43, sub: 
Deferment of Michaelmas, and Rad, Gen R. J. 
Marshall to Gen Chamberlin, 28 Dec 43, both in 
GHQ SWPA G-3 Jnl, 29 Dec 43, 

in order to reach Saidor on D Day, Mar- 
tin concluded that loading would have 
to start on 30 December, for the task 
force assembly area on Goodenough lay 
eighteen miles from the embarkation 
point. He ordered his force to move to 
the embarkation point at once. Move- 
ment and loading continued night and 
day, usually in rain which turned the 
roads into mud. On 30 December Martin 
received word officially that D Day 
would definitely be 2 January. 

APD's and LCI's took aboard only 
troops, individual equipment, and in- 
dividual or squad weapons. Heavy equip- 
ment, vehicles, motor-drawn weapons, 
bulk supplies, and some troops went 
aboard the LST's. Martin was forced to 
make a last-minute change in embarka- 
tion plans on 30 December when he 
found there would be nine APD's in- 
stead of the ten he had expected. The 
surplus infantrymen were ordered aboard 
an LCI, but since the units involved 
were then moving to the embarkation 
point some did not receive word of the 
change until they had reached the beach. 

The difficulties were all overborne, 
and by 0830, 3 1 December, the six LST's 
had departed Goodenough. Their slots 
at the beach were promptly taken by the 
six that were to bring heavy equipment 
to Saidor on D plus 1. The LCI's left 
Goodenough in midafternoon, and the 
fast APD's completed loading at 1700. 
There had not been opportunity for a 
dress rehearsal, but the LCP(R)'s of the 
APD's practiced landing formations at 
Goodenough Bay. 50 

50 The 126th Regimental Combat Team had re- 
ceived six weeks of amphibious training in Aus- 
tralia and three weeks training in LST's, LCI's, and 
APD's at Milne Bay. 



For the soldiers and sailors aboard the 
APD's New Year's Eve passed quietly. 
Martin, the APD captains, Colonel Brad- 
ley, and the battalion commanders con- 
ferred aboard the destroyer Stringham 
and made minor last-minutes changes in 
landing plans. Some of the ships showed 
moving pictures. At 0600 on New Year's 
Day, 1944, the APD's sailed. They put in 
at Oro Bay en route, where they were 
joined by Admiral Barbey in his flag- 
ship, the destroyer Conyngham, to which 
General Martin transferred his command 
post afloat. During the ships' approach to 
Saidor on 1 January, sixty B-24's and 
forty-eight B-25's hit the Japanese instal- 
lations with 218 tons of demolition 

Barbey's final run through the straits 
to Saidor was unexciting. The early part 
of the night of 1-2 January was clear, 
with a quarter moon shining on the 
ships. After midnight the sky became 
overcast and rain fell. 

Seizing Saidor 

When the ships and landing craft hove 
to in Dekays Bay before sunrise of 2 
January, heavy overcast and rain ob- 
scured the shore. Admiral Barbey post- 
poned H Hour from 0650 to 0705 to 
provide more light for naval gunfire, 
loading and assembly of boats, and 
identification of beaches. There fol- 
lowed another delay of twenty minutes 
while LCP(R)'s formed up. The destroy- 
ers and rocket LCI's fired the scheduled 
1,725 5-inch shells and 624 4.5-inch 
rockets at the beaches and inland areas. 
Troops aboard ship, one thousand yards 
offshore, felt the concussion of the ex- 

The assault waves meanwhile had 

boated and assembled, and were churn- 
ing toward Red, White, and Blue 
Beaches. First craft touched down at 
0725, and during the next seventeen 
minutes the four waves of thirty-six 
LCP(R)'s landed 1,440 troops. There was 
no opposition from the enemy. The six- 
teen LCI's, organized in three waves, 
grounded and put ashore more than 
3,000 troops. 

Each LST had towed an LCM of the 
2d Engineer Special Brigade. The LST's 
cast loose their tows on arrival offshore, 
and three LCM's sailed to the beaches 
with the last wave of small craft. Thirty 
minutes before the LST's were sched- 
uled to beach, an angledozer clanked out 
of each LCM and at once set to work 
grading landing points and beach exits 
to use in unloading the LST's. When the 
six LST's beached at about 0800, landing 
points and beach exits made of gravel 
and wire mesh were ready. This per- 
formance, plus the efficiency of the shore 
party, which Admiral Barbey praised 
highly, enabled cargo to come ashore in 
record time. Each LST rapidly unloaded 
three hundred tons of bulk supplies and 
two hundred tons of vehicles and equip- 
ment. By 1 140 all LST's had unloaded 
and retracted. The bad weather delayed 
the scheduled air bombing, but later in 
the morning B-24's, B-25's, and A-20's 
bombed the Saidoi; airstrip and the high 
ground inland. 51 

51 There was one hilch in the air plans. Martin 
did not receive the ist Air Task Force's support 
plan until after he had left Goodenough and was 
aboard ship. There he discovered that one alternate 
target lay on the American side of the bomb safety 
line. As the ships were under radio silence he could 
not notify air headquarters. The air liaison party, 
after landing, radioed the necessary information di- 
rectly to the bombers before any American casual- 
ties were incurred. 



Mio Motor Carriage Mounting ^-Inch Gun on a rough and stony beach near 
Saidor, 2 January 1944- 

When they reached shore the rifle bat- 
talions began to push inland while the 
artillery established itself and the shore 
party moved supplies off the beach. Jap- 
anese resistance was limited to a few rifle 
shots. General Martin reported that only 
15 enemy soldiers had been near the 
beaches at the time of the landing, and 
1 1 of these were killed by the bombard- 
ments and by soldiers of the 126th. 
Saidor had a normal garrison of about 
50, and on 2 January some 120-150 
transients were present. All these 
promptly took to the hills. American 
casualties on D Day numbered 1 soldier 
killed and 5 wounded and 2 sailors 
drowned at Blue Beach. Forces ashore 
numbered 6,779: 52 6,602 Army ground 

52 This figure comes from the Michaelmas Task 
Force report. CTF 76's report gives 7,200. 

troops, 129 from Army Air Forces units, 
and 48 sailors. 

Admiral Crutchley's task force had 
performed its usual mission of covering 
the invasion against Japanese warships 
from Rabaul, but none appeared. 
Thirty-nine fighters and twenty-four 
bombers of the 4th Air Army were based 
at Wewak but were unable to launch an 
attack until 1600. By then Barbey's ships 
were well out to sea, Martin's soldiers 
had dispersed their supplies, and little 
damage was done. 

So ended the first day at Saidor. The 
speedy efficiency of Saidor operations, 
when compared, for example, with the 
Kiriwina invasion of the previous year, 
bears witness to the Southwest Pacific's 
improvement in amphibious technique. 
Yet the fact that there are flaws in even 



the best-executed operations was demon- 
strated the next morning. The Michael- 
mas Task Force expected six LST's (each 
towing an engineer LCM) carrying the 
1 2 ist Field Artillery Battalion and A and 
D Batteries, 743d Coast Artillery Gun 
Battalion (Antiaircraft), to arrive at 
0700, 3 January, which would be after 
daylight. When three vessels came dimly 
in view about a hundred yards from the 
north shore at 0510 and failed to identify 
themselves, the shore defenses opened 
fire. The vessels withdrew. After daylight 
they returned and were correctly identi- 
fied as three of the six expected LST's. 

Thereafter shipments of troops and 
supplies on the LST's were uneventful. 
The 808th Engineer Aviation Battalion 
arrived on 6 January; the larger part of 
the 128th Regimental Combat Team 
came in on 16 and 19 January in response 
to General Martin's request for more 
troops; the 863d Engineer Aviation Bat- 
talion was landed on 29 January. By 10 
February, when General Krueger de- 
clared Dexterity over, and GHQ an- 
nounced that U.S. Army Services of Sup- 
ply, Southwest Pacific, would take over 
supplying Saidor, Arawe, and Gloucester 
on 1 March, the Saidor garrison num- 
bered 14,979 in addition to a small naval 
detachment. Forty men had been killed, 
1 1 1 wounded; if) were missing. 

Base Development 

Construction missions assigned to the 
engineers included building or installing 
an airfield, roads, fuel storage tanks, 
docks, jetties, a PT boat base, and a hos- 
pital. Work on the airfield started 
promptly and in itself was not difficult, 
since the prewar field was in fair condi- 
tion. The amphibian engineers unloaded 

ships and built roads, but continuous 
rainfall hindered their work and General 
Martin occasionally diverted the aviation 
engineers to assist the amphibians in 
their work. The Americans were assisted 
in all phases of construction work by 
native labor. A detachment of the Aus- 
tralian New Guinea Administrative Unit 
consisting of one officer, several enlisted 
men, and eleven native policemen had 
landed on 2 January to supervise the em- 
ployment of New Guinea natives. Seven 
days later 100 native workers came up 
from Lae, and by 10 February there were 
13 policemen, 200 Lae natives, and 406 
local workers at Saidor. C-47's landed on 
the airfield on D plus 9, and by 10 Feb- 
ruary it was almost ready to receive war- 

Junction With the Australians 

In the days immediately following D 
Day General Martin disposed one battal- 
ion in defensive positions on each flank, 
with the third patrolling in the moun- 
tains between the flanks. There were 
some fourteen miles of coast line be- 
tween the flanks which were held by the 
coast artillery, supported by other units. 

The Michaelmas Task Force was 
hardly ashore when General Krueger 
warned that, as the Japanese troops in 
the vicinity of Sio were preparing to 
move west to Madang, an attack against 
Saidor was to be expected. These warn- 
ings were repeated on 7 and 9 January. 
General Martin asked for more troops 
on 10 January and General Krueger sent 
him the 1st and 3d Battalion Combat 
Teams of the 128th Regimental Combat 

Meanwhile patrols went to the east, 
west, and south. There were occasional 



brushes with scattered Japanese patrols, 
but no pitched battles. January ended 
without an attack. The Japanese were 
known to be advancing west, but they 
had not yet touched Saidor. 

Almost from the very outset General 
Martin had urged an advance to the east 
to hem in the Japanese between his 
forces and the advancing Australians, 
but, partly because it apparently did not 
wish to commit additional troops, and 
partly because of garbles in the transmis- 
sion of messages, Alamo headquarters 
did not at once accede to Martin's de- 
sires. 53 

Doubts regarding Japanese intentions 
were dispelled on 6 February when from 
a newly established observation post in 
the Finisterres American soldiers saw 
large numbers of Japanese troops march- 
ing along an inland trail that ran south 
of Saidor through the mountains and 
foothills. It was concluded that the Japa- 
nese were bypassing Saidor. 

The conclusion was correct. In late 
December General Adachi, concerned 
over the state of things to the east, had 
flown from Madang to 51st Division 
headquarters at Kiari. He received word 
of the American landing at Saidor just 
before he went overland to the 20th Di- 
vision at Sio. Opinion at General Ima- 
mura's Rabaul headquarters was divided 
over the best course of action. Some staff 
officers argued that the 20th and 51st Di- 
visions should attack Saidor. Others 
counseled that they slip peacefully past 
Saidor over an inland route and proceed 

03 Gen Martin, ist Ind, 2 Nov 53, to Ltr, Gen 
Smith, Chief of Mil Hist, to Gen Martin, 6 Oct 53, 
no sub, OCMH. General Martin kindly attached to 
his indorsement several letters and papers and took 
the trouble to prepare an excellent narrative which 
clarified many obscure points. 

to Madang to join with the remainder of 
the 1 8th Army to defend Wewak. Aban- 
doning the attempt to hold the shores of 
the straits, Imamura decided in favor of 
bypassing Saidor. He sent orders to that 
effect to Adachi at Sio. 

Adachi placed General Nakano of the 
51st Division in charge of the retreat and 
directed the 41st Division to move from 
Wewak to defend Madang. Adachi left 
Sio by submarine "in a troubled state of 
mind because he would again have to 
force the two divisions to go through dif- 
ficulties." 54 He later ordered General 
Nakai to send eight companies out of 
the Ramu Valley to Bogadjim. They 
were to advance down to the coast to 
harass Saidor while Nakano's force re- 

Nakano, who first directed the 20th 
Division to retreat along the coast while 
the 51st Division and naval units moved 
inland, eventually decided to avoid 
enemy opposition by sending the whole 
force through the Finisterres. 

The retreat began promptly. Sio was 
abandoned to the 9th Australian Di- 
vision and the Japanese moved up the 
coast, then headed inland east of Saidor. 
The retreat to Madang, almost two hun- 
dred miles away by the coastal route, was 
another of the terrible Japanese marches 
in New Guinea. The troops struggled 
through jungles, across rivers, and over 
the awesome cliffs and mountains of the 
Finisterres. Fatigue, straggling, disease, 
and starvation characterized the retreat. 
"The men were no longer able to care 
for themselves and walked step after step 
looking ahead only a meter to see where 

54 18th Army Operations, II, Japanese Monogr No. 
43 (OCMH), 132. 



they were going." 55 The two divisions 
had totaled twenty thousand in Decem- 
ber 1943; only ten thousand wearily 
entered Madang in mid-February. 

Yet that the ten thousand made such 
a trip and that the Japanese could make 
such marches in retreat and in the ad- 
vance are tribute and testimony to the 
patient fortitude and iron resolution of 
the Japanese soldier. They clearly illus- 
trate that despite his baggy uniforms and 
bombastic phrases he was a formidable 

After the fall of Sio the 5th Aus- 
tralian Division relieved the gth and ad- 
vanced up the coast. Its advance patrols 
made contact with those of the Michael- 
mas Task Force on the Yaut River about 
fourteen miles southeast of Saidor on 10 
February 1944. 

Because permission to move east was 
received too late, Martin could not block 
the Japanese in that direction. And the 
escape route to the south ran up and 
down such steep ravines and slopes that 
no heavy weapons could be carried there, 
and the Americans could not block that 
route either. General Martin decided to 
attack to the west. The move, executed 
by elements of the 1st and 3d Battalions, 
126th Infantry, began at once. 

General Gill and his staff reached 
Saidor on 18 February to assume com- 
mand, and continued the westward 
move. By 24 February patrols of the 3d 
Battalion, 126th, had reached Biliau at 
Cape Iris, about twelve land miles from 
Saidor. On 5 March the 126th Infantry 
(less the 2d Battalion), the 121st Field 
Artillery Battalion, and B Battery, 120th 

55 18th Army Operations, II, Japanese Monogr No. 
42 (OCMH), 180. 

Field Artillery Battalion, disembarked 
from engineer landing craft at Yalau 
Plantation, twenty miles farther on. By 
now the 7th Australian Division had 
broken out of the Ramu Valley and Gen- 
eral Nakai was retreating toward Ma- 
dang. Patrols of the 32d U.S. and the 7th 
Australian Divisions made contact at Kul 
between the Sa and Kambara Rivers, 
about eight miles beyond Yalau Planta- 
tion, on 21 March. The Australians went 
on to take Bogadjim on 13 April. 

Meanwhile Imperial General Head- 
quarters had transferred the 18th Army 
and the 4th Air Army out of the 8th Area 
A rmy and assigned them to control of the 
2d Area Army to pull Adachi's troops out 
of Madang and west to Wewak, Aitape, 
and Hollandia. Adachi's troops started 
west again, evacuating Madang just be- 
fore the Australians entered from the 
east on 24 April. 50 So the large-scale at- 
tack on Madang envisaged in Elkton 
III never came off. Saidor proved to be 
an effective and economical substitute. 

Securing of the major objectives of 
Operation II of Cartwheel was com- 
pleted by the seizure of Saidor, and sub- 
sequent operations on the Huon Penin- 
sula were anticlimactic strategically, 
however bitter and tragic they were for 
those who fought and died in them. The 
Saidor landing completed the seizure of 
the Markham-Ramu trough and the 
Huon Peninsula for the Allies and ob- 
tained one more airfield to support oper- 
ations against the Admiralties and enemy 
bases to the west. 

56 For the iSth Army's operations subsequent to 
the period embraced by this volume, see Smith, 
The Approach to the Philippines, Chs. V-VIII. The 
new boundary between the Sth Area and the 2d 
Area Armies was set at longitude 147° east. 



Strictly speaking, Saidor was the last 
invasion of the Cartwheel operations. 
With it General MacArthur fulfilled the 
provisions of the Joint Chiefs' orders of 
28 March 1943. But he and Admiral 
Halsey were not yet finished with the 
Japanese in the Southeast Area. By the 
end of 1943 Rabaul had not yet been 
completely neutralized, and before the 

approach to the Philippines could begin 
there remained a set of subsidiary, 
transitional operations to be accom- 
plished. These, which the Joint Chiefs 
and MacArthur had discussed earlier in 
1 943> would complete the encirclement 
of Rabaul and would provide a naval 
base to substitute for Rabaul in the drive 
to the Philippines. 


Expanding Into the Bismarck 

General Plans 

Further operations in the Bismarck 
Archipelago had been contemplated for 
nearly two years. The Joint Chiefs' di- 
rective which launched the campaigns 
against Rabaul in 1942 had authorized 
operations to follow Arawe and Cape 
Gloucester, and MacArthur's early plans 
called for the capture of Kavieng on 
northern New Ireland and of Manus in 
the Admiralty Islands as well as of 
Rabaul. 1 Further, when the Joint Chiefs 
were deciding to bypass Rabaul, Gen- 
eral Marshall suggested that Cartwheel 
be extended to include seizure of Ka- 
vieng, Manus, and Wewak. MacArthur 
was less than enthusiastic about Wewak, 
which was a major enemy base. His plan 
for the drive to the Philippines, Reno 
III, called for the invasion of Hansa Bay 
on 1 February 1944, of Kavieng by the 
South Pacific on 1 March 1944, and of 
the Admiralties on 1 March 1944. 2 

Responsibility for base construction at 
Kavieng and at Seeadler Harbour at 

1 See above, pp. Qi^i. I 

2 See also GHQ SWPA Warning Instns 3, 23 Nov 
43, in Alamo Anchorage Jnl 1, 23 Nov 43-12 Feb 

Manus was to be Admiral Halsey's, and 
he began planning these bases in No- 
vember 1943. Kavieng was supposed to 
become a minor fleet base, a PT base, 
and a major air base with six airfields. 
Manus would serve as an air base (two 
airfields and a seaplane base) while 
Seeadler Harbour would be developed 
into a major fleet base whose complete 
repair facilities would include drydocks, 
and a main naval supply base. It would 
serve Admiral Nimitz' naval forces as 
well as the Seventh Fleet. 3 

Halsey, who conferred with Mac- 
Arthur in Brisbane in late 1943 before 
departing on a trip to Hawaii and the 
continental United States, opposed the 
seizure of Kavieng. He wished to apply 
the bypass technique and seize Emirau in 
the Saint Matthias Islands, about ninety 
miles northwest of Kavieng, for this 
group had never been taken by the Jap- 
anese. Kavieng, on the other hand, was a 
major air and naval base and was re- 
ported to be strongly defended. In De- 
cember MacArthur told members of 
Halsey's staff that an attack against Emi- 

3 File on Manus-Kavieng base development in 
GHQ SWPA G-3 Jnl, 5 Nov 43. 



rau or Kavieng would serve equally well 
in the isolation of Rabaul. 4 

Halsey spent four days with Nimitz at 
Pearl Harbor and then, in early January, 
flew to San Francisco where he and 
Nimitz conferred with Admiral King. 
Here, and later in Washington, the South 
Pacific commander made known his 
views on Kavieng and Emirau. 5 

Halsey was not able to carry his point 
at this time. He did however discuss tim- 
ing and the question of naval support 
for Manus and Kavieng. These were im- 
portant because by now the Central 
Pacific offensives were well under way. 
Nimitz' forces, having invaded the Gil- 
berts in November 1943, were planning 
their initial move into the Marshalls 
(Kwajalein and Roi-Namur) in late 
January. Kavieng, almost four hundred 
miles from Cape Torokina, lay beyond 
fighter-plane range of Halsey's most ad- 
vanced air base. Thus aircraft carriers 
would have to provide cover for the in- 
vasion forces, and Admiral Nimitz 
agreed to furnish them. General Mac- 
Arthur wanted carriers to cover the in- 
vasion of Manus as well, in case bad 
weather kept the Fifth Air Force planes 
grounded in New Guinea and at Cape 
Gloucester. Nimitz pointed out, how- 
ever, that such weather could also affect 
carrier operations. 

Admiral Carney, Halsey's chief of staff, 
had visited Pearl Harbor in December 

1 Memo, SJC [Chamberlin] for Jnl, 21 Dec 43, 
sub: Conf at GHQ, 20 Dec 43, in GHQ SWPA G-g 
Jnl, 21 Dec 43. 

1 Halsey and Bryan, Admiral Halsey's Story, pp. 
186-87; Ernest J. King and Walter Muir Whitehall, 
Fleet Admiral King: A Naval Record (New York: 
W. W. Norton & Company, Inc., 1952), pp. 533-34. 

COMSOPAC, 7 Jan 44, CM-IN 8330. 

and reported that the ships for Kavieng 
w r ould not be available until 1 May. 7 
This would also put off the Admiralties 
operation. 8 But Admiral Nimitz then 
suggested that by delaying his second 
Marshalls invasion (Eniwetok) until 1 
May he could provide support for Manus 
and Kavieng about 1 April. MacArthur 
was ready and willing to invade Manus 
and Kavieng in March before moving to 
Hansa Bay, but the Joint Chiefs ordered 
Nimitz to deliver a strong carrier strike 
against Truk during March. No direct 
naval supporting forces could be avail- 
able for Manus and Kavieng until April. 8 
Nimitz proposed that representatives of 
all the Pacific areas meet in Pearl Har- 
bor to settle details of co-ordination and 
timing. 10 

The command question came up again 
in January when Marshall asked Mac- 
Arthur's opinion on a draft directive for 
the next operations. The draft, Marshall 
told him, had received the approval of 
General Kenney, who was also in Wash- 
ington. Except for Kavieng it did not 
specify any particular localities to be at- 
tacked but authorized advances into the 
Bismarck Archipelago preparatory to the 
drive to the Philippines. South Pacific 
forces attacking Kavieng were to be 
placed under MacArthur's "general di- 

7 Memo, Carney for Halsey, 12 Dec 43, sub: 
CINCPOA-SOPAC Stf Conf, 9-12 Dec 43, in GHQ 
SWPA G-3 Jnl, 21 Dec 43. 

8 Memo, B F [Brig Gen Bonner Fellers, G-3 Sec 
GHQ SWPA], no addressee, 22 Dec 43, sub: Conf 
G-3 Ping Sec, in GHQ SWPA G-3 Jnl, 22 Dec 43. 

•Rad, CINCPAC to COMINCH, 22 Dec 43, in 
GHQ SWPA G-3 Jnl, 24 Dec 43; Rad, Halsey to 
MacArthur, 5 Jan 44, and Rad, MacArthur to Mar- 
shall and Halsey, 6 Jan 44, and Rad, COMSOPAC 
to COMSOPAC Administration, 9 Jan 44, all in 
Marshall's IN Log. 

10 Rad, CINCPAC to CINCSWPA, CNO, and 
COMSOPAC, 7 Jan 44, CM-IN 8330. 



rection," and Nimitz was ordered to pro- 
vide fleet support and more assault ship- 
ping for Manus and Kavieng after the 
approaching conference at Pearl Har- 
bor. 11 

MacArthur objected strenuously. After 
reviewing the course of Cartwheel 
operations, which took place along two 
axes and for which, therefore, "loose co- 
ordination" sufficed, he argued that in 
the Bismarck Archipelago the South and 
Southwest Pacific forces would be con- 
verging in a fairly restricted area. South 
Pacific forces alone could not capture 
Kavieng, and elements of the forces 
might have to be mingled. Constant, 
complete co-ordination of air and surface 
units would be required. Unity of com- 
mand, vested in himself, should be ap- 
plied, urged -MacArthur, with the South 
Pacific forces under Halsey's direct com- 
mand. And, finally, the Joint Chiefs 
rather than Nimitz should determine the 
extent of fleet support and additional 
assault shipping. 12 

In their orders for the extension into 
the Bismarck Archipelago, dated 23 and 
24 January, the Joint Chiefs acceded to 
MacArthur's suggestions on fleet support 
in a left-handed way. They directed 
Nimitz to provide fleet support and cover 
for the Manus-Kavieng invasions under 
his direct command, and to attach more 
warships and assault shipping to Mac- 
Arthur's and Halsey's forces. The exact 
amounts were to be determined at the 
forthcoming Pearl Harbor conference, 
which would then forward recommenda- 
tions to Washington for approval. 

11 Rad 3097, Marshall to MacArthur, 17 Jan 44, 
in Marshall's OUT Log. 

"Rad, MacArthur to Marshall, 19 Jan 44, in 
Marshall's IN Log. 

Control over South Pacific forces re- 
mained the same as for Cartwheel. 
Halsey was in direct command under 
MacArthur's direction. 13 

The conference at Pearl Harbor con- 
vened on 27 January. Halsey, flying out 
from Washington, had been grounded by 
bad weather in Fort Worth, Texas, and 
again in San Francisco, and so was not 
present. Carney, whom he had au- 
thorized to make preliminary arrange- 
ments with MacArthur, represented him, 
as did General Harmon. Representing 
MacArthur were Sutherland, Kenney, 
and Kinkaid. Nimitz, Rear Adm. Forrest 
P. Sherman, and others spoke for the 
Central Pacific. 

Sutherland made it quite clear that 
MacArthur now definitely wanted the 
South Pacific to capture Kavieng for use 
as an air base, not Emirau. Halsey's pro- 
posal was shelved for the time being. 

Besides discussing operations in the 
Bismarck Archipelago, the conference 
covered a wide range of topics— the value 
of the Marianas, B-29's, the possibility of 
bypassing Truk, and the comparative 
merits of the Central and Southwest 
Pacific routes to the Philippines. All 
agreed that whether Truk was bypassed 
or not, Seeadler Harbour was essential 
as a fleet base for the approach to the 

Nimitz proposed to give long-range 
support to the Manus-Kavieng invasions 
with a two-day strike against Truk by 
the main body of the Pacific Fleet start- 
ing about 26 March. In addition he 
agreed to send two divisions of fast car- 

" JCS 679, 24 Jan 44, title: Dirs for Seizure of 
Control of Bismarck Archipelago; Rads, JCS to 
CINCPAC and CINCSWPA, 23 Jan 44, with JCS 



riers to operate under Halsey's command 
during the Manus-Kavieng invasions, 
while other carrier divisions and fast 
battleships operated in covering posi- 
tions. 14 

These were large forces indeed. As 
originally planned the Bismarck opera- 
tions would have been extensive. In ad- 
dition to the naval forces, Halsey 
planned to use all his land-based air- 
craft and two divisions in assault, with 
one in reserve. However, not one of the 
operations approved by the Joint Chiefs 
and MacArthur was carried out accord- 
ing to the original plan. 

Reducing Rabaul and Kavieng 

All during the invasions of Arawe, 
Cape Gloucester, and Saidor, and during 
the discussions over the Bismarck Archi- 
pelago operations, the Solomons air com- 
mand had been putting forth a maxi- 
mum effort to reduce Rabaul. Comple- 
tion of the Torokina fighter strip at 

11 These forces were to include 3 aircraft carriers, 
3 light carriers, 7 cruisers, and 18 destroyers. In 
addition 4 old battleships, 7 cruisers, 4 escort car- 
riers, 48 destroyers, 30 destroyer-escorts, 1 command 
ship (AGC), 19 transports, 3 LSD's, 5 minesweepers, 
36 LST's, and 36 LCI's would be assigned to 
Halsey's Third Fleet for Kavieng, while for Manus 
the Seventh Fleet was to receive 3 light cruisers, 4 
escort carriers, 35 destroyers, 8 patrol frigates, 1 
AGC, 1 transport, 1 cargo ship, 2 minesweepers, ■ 
LSD, 13 APD's, 30 LST's, 30 LCI's, 70 LCT's, and 
30 submarines. Halsey and Bryan, Admiral Halsey's 
Story, p. 188; Craven and Gate, The Pacific: 
Guadalcanal to Saipan, pp. 551-52; Kenney, Gen- 
eral Kenney Reports, p. 346; Smith, The Approach 
to the Philippines, pp. 7-8; Halsey, Narrative Ac- 
count of the South Pacific Campaign, OCMH; Rad, 
CINCPAC to COMINCH-CNO, 39 Jan 44, in GHQ 
SWPA G-3 Jnl, 30 Jan 44; Ltr, CINCPOA to 
COMINCH, 30 Jan 44, sub: Assignment Naval 
Forces and Assault Shipg to Third and Seventh 
Fits for Opns Against Bismarck Archipelago, ABC 
384 Pac (17 Jan 43) Sec 3-A. 

Empress Augusta Bay, Bougainville, was 
a major step forward, for now New 
Georgia- and Guadalcanal-based bomb- 
ers could have fighter escort in their 
attacks. But by the end of 1943 it was 
clear that high-level bombing would not 
suffice to neutralize Rabaul. Obviously, 
success depended on completion of the 
bomber strips by the Piva River (Piva 
Uncle and Piva Yoke). 15 

Piva Uncle, eight thousand feet by 
three hundred feet, was ready as a stag- 
ing field on 30 December 1943. On 5 
January 1944 SBD's and TBF's from 
Munda staged through to attack Rabaul, 
but by noon, when the bombers arrived 
over the target, Rabaul was as usual 
blanketed by heavy clouds. A similar at- 
tack two days later met the same diffi- 
culties, but on 9 January Piva Yoke was 
ready and from then on bombers could 
be permanently based at the Bougain- 
ville fields and could reach Rabaul in the 
morning, before it was covered by clouds. 

Thereafter during January TBF's, 
SBD's, B-25's, and B-24's struck regu- 
larly at Rabaul. The Japanese lost many 
planes but occasionally received rein- 
forcements from Truk, and continued to 
resist with fighter interception and anti- 
aircraft fire. ". . . the skies overhead 

1S Unless otherwise indicated this section is based 
on Craven and Cate, The Pacific: Guadalcanal to 
Saipan, pp. 350-56; Morison, Breaking the Bis- 
marcks Barrier, pp. 337-66, 392-410; Samuel Eliot 
Morison, History of United Slates Naval Operations 
in World War II, VII, Aleutians, Gilberts and Mar- 
shalls: June 1942— April 1044 (Boston: Little, Brown 
and Company, 1951), 330; Building the Navy's 
Bases in World War II, II, 268-74; USSBS, The 
Allied Campaign Against Rabaul; 8th Area Army 
Operations, Japanese Monogr No. 1 10 (OCMH), p. 
123; Southeast Area Naval Operations, III, Japa- 
nese Monogr No. 50 (OCMH), 6, 58-63. 



Japanese Ships Burning at Raba 

were the scene of continuous annihila- 
tion battles. . . ." 16 

By the end of January heavy bombers 
had flown 263 sorties over Rabaul; 
B-25's, 180 sorties; SBD's, 368; TBF's, 
227; fighters, 1,850. Losses totaled 8 
B-24's, 14 B-25*s, 8 SBD's, 5 TBF's, 19 
P-38's, 37 F4U's, 5 F6F's, and 6 New 
Zealand P-40's. 

Damage to Japanese equipment and 
weapons on the ground was relatively 
light, for in late November the enemy 
had begun the prodigious task of digging 
every possible item underground in Ra- 
baul's volcanic rock, a task that was well 
along by January. But all buildings were 
knocked flat. Ships and grounded planes 

"Southeast Area Naval Operations, III, Japanese 
Monogr No. 50 (OCMH), 58. 

were especially vulnerable to low-level 
bombing and dive-bombing. By Febru- 
ary 1944 the Allies had won a signal vic- 
tory; Japanese surface ships stopped 
using the harbor. 

During the same period Kavieng re- 
ceived increased attention from both 
Allies and Japanese. Halsey, expecting 
to assault the base eventually, wanted to 
reduce Kavieng to help cut the Japanese 
lines of communication from rear bases 
to Rabaul. The Japanese, well aware of 
the threat to Rabaul, decided to 
strengthen Kavieng and the Admiralties 
to help protect Rabaul. 

In October Imamura had sent the 
230th Infantry of the 38th Division from 
Rabaul to New Ireland. Next month he 
sent an emissary to Tokyo to ask for one 



more division. Imperial Headquarters 
responded by sending the ist Independ- 
ent Mixed Regiment to New Ireland. It 
reached its destination in late 1943 and 
early 1944. Imamura placed it, together 
with the 230th Infantry, under Maj. 
Gen. Takeo Ito, infantry group com- 
mander of the 38th Division. Ito's 
soldiers and the 14th Naval Base Force 
were responsible for defense of New 

In December Halsey set a trap and 
ordered Buka bombarded to lure Japa- 
nese planes and ships away from Ka- 
vieng. Admiral Sherman, lying east of 
Kavieng with the carriers Bunker Hill 
and Monterey plus escorts, was then to 
strike at Kavieng in the hope of catch- 
ing troopships and warships in the har- 
bor. Before dawn on Christmas morning 
Sherman launched eighty-six planes, 
which bombed Kavieng at 0745 and were 
back aboard their carriers by 1015. But 
the results were disappointing as there 
were almost no ships in the harbor. 

On New Year's day Sherman delivered 
another strike from 220 miles east of 
Kavieng. Outside the harbor his planes 
caught some of the ships that had just 
unloaded part of the ist Independent 
Mixed Regiment but the Japanese air 
cover of forty-two planes prevented the 
ships from suffering damage. Sherman 
struck Kavieng again three days later, 
again without doing much damage; no 
ships were present and the Japanese 
planes were out against Cape Gloucester. 

In February the Fifth Air Force, using 
Finschhafen as a fighter base and Cape 
Gloucester as an emergency field, began 
to attack Kavieng with the aim of soften- 
ing it before the projected invasion, cut- 
ting the line of communications to Ra- 

baul, and supporting the South Pacific's 
invasion of the Green Islands. On the 
11th forty-eight B-24's with P-38 escorts 
caught Kavieng's planes on the ground, 
and the next two days saw similar attacks. 

During the first two weeks of Febru- 
ary Rabaul's defenses grew obviously 
weaker as the Air Command, Solomons, 
maintained the intensity of its attack. 17 
There were few attempts to intercept un- 
til 19 February. On that date twenty- 
eight SBD's, twenty-three TBF's, and 
sixty-eight fighters, finding no ships in 
the harbor, put bombs and rockets on 
Lakunai airfield. Twenty B-24's with 
thirty-five escorting fighters bombed 
from high altitudes. About fifty Japa- 
nese fighters attempted to break up the 
attack without success. This was the last 
attempted interception. Thereafter at- 
tacking Rabaul became a milk run. 
Allied pilots encountered antiaircraft 
fire but no planes. Rabaul no longer 
could threaten any Allied advance except 
one directed against itself. 

Rabaul's impotence was of course 
largely brought about by the South and 
Southwest Pacific air and naval cam- 
paigns that had been under way for so 
long, but it was partly brought about by 
Admiral Nimitz' naval forces. The Cen- 
tral Pacific had invaded Kwajalein and 
Roi-Namur on 31 January and seized 
them so rapidly that the reserve and gar- 
rison forces did not have to be com- 
mitted. When the Joint Chiefs told 
Nimitz that they were willing to delay 
the Manus-Kavieng invasions in order to 
proceed directly to Eniwetok, using the 
uncommitted troops, Nimitz decided to 

11 From 1 through ig February fighters flew 1,579 
sorties; B-24's, 256; B— 25's, 263; TBF's, 244, and 
SBD's 573. 



go there as quickly as possible. Accord- 
ingly he invaded Eniwetok on 17 Feb- 
ruary. In support of this move the main 
body of the Pacific Fleet, commanded by 
Vice Adm. Raymond A. Spruance, at- 
tacked Truk on 16 and 17 February, 
over one month ahead of schedule. Spru- 
ance's strike was an outstanding success. 
The Combined Fleet had already es- 
caped toward home waters, but Spru- 
ance's pilots destroyed or damaged 250- 
275 planes as well as thousands of tons 
of shipping. Admiral Koga, thus almost 
bereft of planes, ordered all naval planes 
out of the Southeast Area at once. ". . . 
Rabaul, compelled to face the enemy 
with ground resources alone and com- 
pletely isolated, was abandoned." 18 

The Allies dropped 20,584 tons of 
bombs on Rabaul throughout the war, 
and fired 383 tons of naval shells after 
Rabaul was reduced to the indignity of 
suffering destroyer and nocturnal PBY 
bombardment in March. Thirty naval 
vessels were sunk, 23 damaged. In addi- 
tion 154 large cargo vessels and 517 
barges were sunk; 70 small cargo vessels 
suffered damage. 19 ". . . The [Japanese] 
Navy lost the pick of its flight personnel 
at Rabaul, a fact which told heavily upon 
subsequent efforts to rebuild our air 
forces." 20 

Rabaul was abandoned only in the 
strategic sense, and it was impotent only 
for offensive action. It could have de- 
fended itself with bloody efficiency had 

" Southeast Area Naval Operations, III, Japanese 
Monogr No. 50 (OCMH), 6. 

19 These and subsequent statistics come from 
USSBS, The Allied Campaign Against Rabaul, pp. 
1 1-3 6 - 

K Outline of Southeast Area Naval Air Operations, 
November 194a— June 1943, Pt. HI, Japanese 
Monogr No, 107 (OCMH), p. 59. 

the Allies attacked. The garrison of New 
Britain numbered almost 98,000 men 
(76,300 in the 8t.h Area Army and 21,570 
in the naval forces). The rugged country 
of Gazelle Peninsula was well suited for 
defense. By the war's end some 350 miles 
of tunnels and caves had been excavated. 
At peak strength Rabaul had 367 anti- 
aircraft guns (of which 73 were destroyed 
by air bombing), ranging in type and 
caliber from 13.2-mm. to 120.7-mm. dual 
purpose. There were 43 coast defense 
guns (1 destroyed) of calibers up to and 
including 150-mm. Of the 475 artillery 
guns and howitzers (37-mm. to 150-mm.), 
none was destroyed by bombing, nor 
were any of the 1,762 machine guns. Ima- 
mura's men also had tanks, mines, 
ditches, caves, bunkers, and concrete pill- 
boxes, as well as rifles, grenades, bayo- 
nets, and ample ammunition. 

Rabaul would not have been as valu- 
able to the Allies as it was to the Japa- 
nese in their southward advance. It 
would have been useful to the Allies 
only in a northward move against Truk 
and the Marianas. Because the Joint 
Chiefs had decided to advance westward, 
and because Seeadler Harbour in the Ad- 
miralties was better than Rabaul's, the 
Japanese fortress was not worth the price 
the Japanese surely would have exacted. 

Seizure of the Green Islands 

Plans and Preparations 

In December 1943 Admiral Halsey's 
planes were bombing Rabaul, his ships 
were patrolling the Solomon Sea, and his 
ground troops in Bougainville were 
either fighting the enemy or consolidat- 
ing positions in anticipation of a fight. 



But this was not enough to satisfy him. 
When he learned that Nimitz' plans, as 
they stood in December, would not per- 
mit the invasion of Manus and Kavieng 
for several months, he decided to seize an 
air base site within fighter range of Ka- 
vieng in the meantime. 21 At a conference 
in Port Moresby on 20 December at- 
tended by MacArthur, Kinkaid, Car- 
ney, Chamberlin, and others, the South 
Pacific representatives proposed that the 
Southwest Pacific attack Manus directly 
while South Pacific forces captured the 
Green Islands, some 37 miles northwest 
of Buka, and established there an airfield 
and PT boat base. Situated 117 miles 
east of Rabaul and 220 miles southeast 
of Kavieng, this circular coral atoll was 
not strongly held. The Japanese used it 
only as a barge staging base between 
Rabaul and Buka. Allied seizure of the 
atoll would put South Pacific fighter 
planes within range of Kavieng, extend 
the range of PT boat patrols as far as 
New Ireland, and cut the Japanese sea- 
borne supply route to Buka. 

MacArthur, deciding for the time 
being against a move to Manus in ad- 
vance of the projected invasion of Hansa 
Bay, approved simultaneous attacks 
against Manus and Kavieng and told the 
South Pacific to go ahead with the plan 

21 Unless otherwise indicated this section is based 
on Craven and Cate, The Pacific: Guadalcanal to 
Saipan, p. 355; Gillespie, The Pacific, pp. 168-95; 
Halsey and Bryan, Admiral Halsey's Story, p. 188; 
Morison, Breaking the Bismarcks Barrier, pp. 412— 
19; Rentz, Bougainville and the Northern Solomons, 
pp. 114—17; Building the Navy's Bases in World 
War II, II, 274—76; Halsey, Narrative Account of 
the South Pacific Campaign; Southeast Area Naval 
Operations, III, Japanese Monogr No. 50 ^OCMH), 

to attack the Green Islands about 1 Feb- 
ruary. 22 

The island group consists of four flat, 
thickly wooded coral atolls which en- 
circle a lagoon. The group is about nine 
miles long from north to south, five miles 
from east to west. Horseshoe-shaped 
Nissan, the main island, provided good 
landing beaches on its west shore inside 
the lagoon, but it was not known 
whether the passage between Barahun 
and Nissan would accommodate landing 
craft. Therefore Halsey sent four PT 
boats from Cape Torokina to examine 
the passage on the night of 10-11 Jan- 
uary. They found seventeen feet of water 
there, or enough to float an LST. 23 

Admiral Halsey, who returned to 
Noumea on 3 February, placed control 
of the operation and responsibility for 
the co-ordination of amphibious plan- 
ning in Admiral Wilkinson's hands on 5 
February. 24 This action confirmed warn- 
ing orders which had been issued in early 

Only destroyer-transports and landing 
craft were assigned to the attack force. 
Command of the landing force was given 
to General Barrowclough of the 3d New 
Zealand Division. * Barrowclough's di- 
vision (less the 8th Brigade Group), the 
976th Antiaircraft Artillery Gun Battal- 
ion of the U.S. Army, a PT base unit, 
communications units, a boat pool, and 
a large naval base unit including an en- 

22 Memo, SJC [Chamberlin] for Jnl, 21 Dec 43, 
sub: Conf at GHQ, 20 Dec 43, in GHQ SWPA G-3 
Jnl, 21 Dec 43. 

23 A fully loaded LST draws 14 feet 114 inches of 
water astern, 9 feet 914 inches when loaded for 
landing operations. 

51 Rad, Comdr Third Fit to CINCSWFA, CINC- 
PAC, and all TF's SOPAC, 5 Feb 44, in GHQ 
SWPA G-3 Jnl, 7 Feb 44. 



tire construction regiment, constituted 
the landing force. Halsey ordered the 
Solomons air command and Ainsworth's 
and Merrill's cruiser task forces to sup- 
port and cover the invasion, and ar- 
ranged with MacArthur for Kenney's air 
forces to deliver the attacks on Kavieng 
during the first fifteen days of February. 25 

As South Pacific headquarters esti- 
mated that Rabaul and Kavieng would 
be virtually neutralized by mid-Febru- 
ary, D Day was set for the 15th. General 
Barrowclough, who had been island com- 
mander at V ella Lavella, moved his head- 
quarters to Guadalcanal in January to be 
near Wilkinson during the planning. 
They decided to send a large reconnais- 
sance party to Green in order to deter- 
mine the strength of the enemy garrison 
and to examine possible airfield sites, 
beaches, and naval base sites, and the 
lagoon tides. The party was to spend 
twenty-four hours ashore. 

Three hundred and twenty-two 
soldiers of the 30th New Zealand Bat- 
talion and twenty-seven American and 
eleven New Zealand hydrographic, air, 
small boat, communications, and intel- 
ligence specialists boarded three APD's 
on 29 January. The destroyer-transports 
hove to west of Barahun about midnight 
and launched landing craft. Two of the 
torpedo boats that had checked the pas- 
sage led the landing craft through to the 

55 Rad, Comdr Third Fit to CINCSWPA, 2a Jan 
44, in GHQ SWPA G-3 Jnl, 23 Jan 44; Rad, Mac- 
Arthur to Comdrs AAF, ANF, Alamo, et al., iy 
Jan 44, in GHQ SWPA G-g Jnl, 27 Jan 44. Rear 
Adm. John F. Shafroth, Halsey's deputy com- 
mander, issued the warning orders on 3 January. 
SOPAC, et al., 3 Jan 44, sub: Warning Orders, in 
GHQ SWPA G-3 Jnl, 6 Jan 44. 

beach. Once ashore the reconnaissance 
party waited for daylight while the 
APD's hauled clear. Guarded by the New 
Zealand soldiers, the specialists set to 
work and gathered their data. They 
found a good airfield site, and estimated 
that the enemy garrison numbered about 
a hundred. The twelve hundred native 
inhabitants proved so friendly and co- 
operative that preliminary naval bom- 
bardment to support the main landing 
was omitted. The specialists were not 
molested, but the enemy fired on one 
landing craft that went to the south part 
of the island where there was an aban- 
doned Roman Catholic mission and 
killed three New Zealanders and one 
American. When Rabaul heard of the 
landing Kusaka sent six bomb-carrying 
fighters to Green. They attacked the 
landing boats but did no damage. 

The APD's reclaimed the New Zea- 
landers and Americans on 31 January 
and returned safely to Guadalcanal. On 
the way back two of the escorting de- 
stroyers sank a Japanese submarine near 
Buka Passage. 

The Japanese Green Islands garrison 
reported it had suffered heavy losses, 
asked for reinforcements, and fled north- 
west in three landing craft to the Feni 
Islands. Kusaka put 123 men aboard a 
submarine on 1 February and sent them 
to Nissan. The submarine hove to off the 
northeast coast about midnight in a sea 
so rough that after 77 men had gone 
ashore, the submarine commander called 
off the operation and returned to Rabaul 
with 46 men still on board. The return 
of the original garrison to Nissan on 5 
February brought total enemy strength 
to 102. 



The Landings 

In the meantime the South Pacific's 
APD's returned from service in the Cape 
Gloucester operation. Shortly before 12 
February the APD's, LST's, LCI's 
LCT's, LCM's, and patrol boats and 
coastal transports of the amphibious 
force took aboard the 5,806-man New 
Zealand-American landing force at Tu- 
lagi, Guadalcanal, the Russells, New 
Georgia, and Vella Lavella. 26 The ships, 
timing their departures so as to meet off 
Bougainville on 14 February, sailed from 
their various ports on the 12th and 13th. 

A Japanese reconnaissance plane 
spotted them west of Bougainville on 14 
February, reported their presence to Ra- 
baul, and kept contact. Admiral Kusaka 
sent thirty-two planes against the ships 
throughout the moonlit night of 14-15 
February. They did no damage to Wil- 
kinson's ships but managed to hit the 
cruiser Saint Louis in Admiral Ains- 
worth's task force, which was operating 
south of Saint George's Channel. Twelve 
Japanese planes were lost. 

The APD's arrived in the transport 
area west of Barahun shortly after 0600 
on 1 5 February and promptly dispatched 
LCVP's toward the passage. Thirty-two 
fighters of the Solomons air command 
were on station overhead. But Kusaka 
did not yield easily. He sent out seven- 

** Eleven destroyers escorted; two aircraft rescue 
boats and two tugs were also in the amphibious 
force. There were 4,242 New Zealanders and 1,564 
Americans in the landing force. 

teen bombers and about fifteen of these 
attacked the landing craft. They scored 
no hits. At the same time Kenney's air- 
men, with four A-20 and seven B-25 
squadrons, delivered a strong blow 
against Kavieng which kept that base 
from attacking the invaders at Green. 

Within two hours all men of the New 
Zealand combat units went ashore on 
Nissan. During the day all ships and 
boats were completely unloaded and 
with the exception of the LCT's, all left 
for the south once they were emptied. 
The LCT's remained as part of the naval 
advanced base. 

Between 15 and 20 February the New 
Zealand infantrymen hunted down and 
killed the Japanese garrison. Ten New 
Zealanders and 3 Americans were killed; 
21 New Zealanders and 3 Americans were 

By 17 March 16,448 men and 43,088 
tons of supplies had been sent to the 
Green Islands. The 2 2d Naval Construc- 
tion Regiment had begun work at once. 
Within two days of the landings a PT 
boat base opened. This extended the 
range of torpedo boat patrols to New 
Ireland and along the entire northeast 
coast of Bougainville. By 4 March a 
5,000-foot fighter field was ready; in late 
March a 6,000-foot bomber field was 
opened. Kavieng now lay within range 
of fighters and light bombers as well as 
heavy bombers from Bougainville. But, 
stripped of its naval planes when Ad- 
miral Koga ordered their withdrawal in 
February, it had already ceased to 
menace the Allies. 


Action in the Admiralties 

The Decision 

First Plans 

By the time Halsey's forces invaded 
the Green Islands, the Southwest Pacific's 
plans for moves to the Admiralties and 
Hansa Bay, which had been started in 
November 1943, were well developed. 1 
On 13 February General MacArthur 
issued operations instructions to the 
South and Southwest Pacific Areas which 
called for these commands to gain con- 
trol of the Bismarck Archipelago and to 
isolate Rabaul by seizing Manus and 
Kavieng about 1 April. 2 To General 
Krueger's Alamo Force, supported by 
Allied Air and Allied Naval Forces, he 
assigned responsibility for the seizure of 
Seeadler Harbour and Manus, as well as 
Hansa Bay. Using naval construction bat- 
talions and Army service units furnished 
by Admiral Halsey, Krueger was to start 
building a major naval base at Seeadler 
Harbour and to develop the Japanese 
airfields at Lorengau on Manus and 
Momote Plantation on Los Negros. Mac- 

1 GHQ SWPA Warning Instns 3, 23 Nov 43, in 
Alamo Anchorage Jnl 1, 23 Nov 43-12 Feb 44. 

2 GHQ SWPA OI 44, 13 Feb 44, in GHQ SWPA 
G-3 Jnl, 13 Feb 44. Unless otherwise indicated this 
chapter is based on [Maj. William C. Frierson] 
The Admiralties: Operations of the 1st Cavalry Di- 
vision {29 February-iS May 1944), AMERICAN 
FORCES IN ACTION (Washington, 1946). 

Arthur also warned Krueger to make 
ready for the drive west along the New 
Guinea coast. 

As in past and future operations, 
Krueger was responsible for the co- 
ordination of plans. But in these orders 
General MacArthur departed from the 
previous practice in his area and adopted 
principles similar to those prevailing in 
the South and Central Pacific areas. He 
specified that the amphibious (naval) 
commander would be in command of all 
assault forces, ships and troops but not 
aircraft, until the landing force was es- 
tablished ashore. Then the amphibious 
commander would pass the command to 
the landing force commander, who 
would become again responsible to 
his normal military superior— General 
Krueger in the case of units assigned to 
the Alamo Force. 3 

The operations, as planned, differed 
from previous ones in another important 
respect. In one general area three sepa- 
rate naval forces would be operating: 
Halsey's, Kinkaid's, and the additional 
forces from Nimitz. Chamberlin there- 
fore suggested that in the event of a 
major naval action command of these 

3 This change was suggested by Chamberlin who 
felt that previous orders had, in this respect, been 
unsound. Note, SJC to CINC, is Feb 44, in GHQ 
SWPA G-3 Jnl, 13 Feb 44. 



forces be vested in Halsey, who would 
be the senior admiral present. 4 This sug- 
gestion was accepted, although for some 
reason it was not followed in similar 
situations at Hollandia and Leyte. 

Forces assigned to General Krueger 
for the Admiralties totaled 45,110 men. 
They included: 

Forces Number of 


Southwest Pacific ground units 25,974 

1st Cavalry Division 
Antiaircraft and coast artillery units 
592d Engineer Boat and Shore Regiment 
1st Marine Amphibian Tractor Bat- 

Engineer, medical, ordnance, quarter- 
master, signal, and naval base units 

Air units: No. 73 Wing, RAAF 2,488 

South Pacific naval construction units.... 9,545 
South Pacific Army service units 7,103 

These were to be concentrated at Oro 
Bay and Cape Cretin. The 6th Division 
was designated as GHQ reserve. 

Hansa Bay was supposed to be in- 
vaded on 26 April by the 24th and g2d 
Divisions. There an air and light naval 
base would assist in the isolation of Ra- 
baul and the Madang-Alexishafen area 
and would support operations west- 
ward. r> 

The Target: Enemy Dispositions 

The Admiralties, lying 200 miles 
northeast of New Guinea, 260 miles west 
of Kavieng, and 200 miles northeast of 
Wewak, were admirably situated to as- 
sist in isolating Rabaul and in support- 
ing the approach to the Philippines. 

They contained two airfields as well as a 
superb harbor. The Japanese had built 
and used the airfields but, possessing 
Rabaul, had never made extensive use of 
Seeadler. This harbor, formed by the 
horseshoe-shaped curvature of the two 
major islands, had a surveyed area 6 miles 
wide, 20 miles long, and 120 feet deep, 
ample for the fleets of World War II. 
Guarding the harbor entrance is a line 
of islets— Koruniat, Ndrilo, Hauwei, 
Pityilu, and other s— which parallel 

Manus' north coast. 6 (Map 19) 

Manus, the largest in the group, is 
separated from Los Negros by a narrow 
strait, Loniu Passage. Forty-nine miles 
from east to west and sixteen miles 
across, Manus is a heavily wooded island 
of volcanic origin. Mangrove swamps 
cover much of the shore line. A range 
of mountains, two thousand to three 
thousand feet in height, extends the east- 
west length. Many of the streams were 
navigable for small boats, and nearly all 
could be forded except when in spate. 
Principal overland routes consisted of 
four native tracks: three ran from the 
north coast over the high country; the 
fourth extended from Lorengau to the 
west part of Manus. 

Los Negros, much smaller than 
Manus, is irregularly shaped and cut by 
several inlets. Papitalai Harbour, an ex- 
tension of Seeadler Harbour, is sepa- 
rated from Hyane Harbour by a low spit 
only fifty yards across. Natives had built 
a skidway over the spit to drag their 
canoes from one harbor to the other. 

1 Ibid. 

3 Memo, Chamberlin for CINC, g Feb 44, sub: 
Outline Plan— Hansa Bay, and Memo, Chamberlin 
for Comdrs, 9 Feb 44, sub: Hansa Bay— SW Pac 
Forces, both in GHQ SWPA G-3 Jnl, 9 Feb 44. 

6 Data in this subsection are derived from [Frier- 
son] The Admiralties; Morison, Breaking the Dis- 
marcks Barrier, pp. 436—37; 8th Area Army Opera- 
tions, Japanese Monogr No. no (OCMH), p. 133; 
Southeast Area Naval Operations, III, Japanese 
Monogr No. 50 (OCMH), 35-36. 

MAP 19 



The center part of Los Negros, in the 
vicinity of Momote, is flat and fertile. 
The swampy region north of the skid- 
way had some coconut plantations. West 
of Momote are three jungled hill masses 
about two hundred feet high. 

The thirteen thousand natives (Mela- 
nesian with some Micronesian admix- 
ture) lived largely in Los Negros and 
eastern Manus. Coconut was the stand- 
ard commercial crop. The natives, who 
sailed their large canoes with skill, also 
dived for trochus shell and pearls. The 
climate— hot and wet— is about the same 
as that of the rest of the region. 

Japanese troops had landed at Loren- 
gau in April 1942 and developed an air- 
field there. The next year they built a 
5,000-foot strip at Momote and improved 
the Lorengau field. Toward the end of 
the year, as the Allies advanced to the 
Markham Valley, the Huon Peninsula, 
and Cape Gloucester, the Japanese be- 
gan using the Admiralties' fields as stag- 
ing points for aircraft flying between Ra- 
baul and Wewak and Hollandia. 

Up to now the garrison had consisted 
of the 51st Transport Regiment, but 
when the Japanese decided to strengthen 
Kavieng they also decided to reinforce 
the Admiralties. Elements of the 14th 
Naval Base Force, the main body of 
which was stationed in New Ireland, 
were sent to Los Negros and Manus. On 
9 December General Imamura directed 
Adachi to send one infantry regiment 
and an artillery battalion from New 
Guinea to be rehabilitated in the Palaus, 
from where they were to proceed to the 
Admiralties. 7 The 66th Infantry reached 
the Palaus safely, but replacements and 

7 The Admiralties remained under Imamura 's 
control after the reorganization in March 1944. 

reinforcing units en route from Japan 
were lost to a U.S. submarine. Then Ima- 
mura organized an infantry and an artil- 
lery battalion in the Palaus out of other 
replacements. These set out for the Ad- 
miralties in January, but their ships were 
so harried by submarines that they 
turned back. Imamura therefore ar- 
ranged with Kusaka for destroyers to 
carry the 2d Battalion of the 1st Inde- 
pendent Mixed Regiment from Kavieng 
to the Admiralties. This movement was 
accomplished on 23-25 January, and at 
the month's end the 1st Battalion, 229th. 
Infantry, was dispatched. Though it suf- 
fered air attack on the way, it arrived 

By 2 February the Japanese garrison 
consisted of the two infantry battalions, 
the <jist Transport Regiment, and 
several naval detachments. In command 
was Col. Yoshio Ezaki, who also led the 
51st Transport Regiment. He disposed 
his main strength on Los Negros to de- 
fend Seeadler Harbour and Momote air- 
field against attack from the north. An 
Allied attack through Hyane Harbour 
was not expected because it was small, 
with so narrow an entrance that landing 
craft would come under fire as they 
passed through. 

"Prepare for Immediate 
Reconnaissance in Force" 

General Kenney's Allied Air Forces 
had prepared elaborate plans for sup- 
porting the Admiralties invasion from 
Dobodura and Nadzab. During January 
and the first two weeks of February his 
planes bombed the Admiralties and Ka- 
vieng, and also continued their attacks 
against the Wewak airfields so as to keep 
them out of action and destroy the 4th 



Air Army's planes. By 6 February Mo- 
mote and Lorengau airfields were un- 
serviceable, and no planes were present. 
Antiaircraft fire had stopped completely, 
not because the guns were destroyed but 
because Colonel Ezaki, to conceal his 
positions from the Allies, had ordered 
his troops neither to fire nor to move 
about in daylight. 

At this time Kenney and Whitehead 
were eagerly seeking methods by which 
the whole advance could be made to 
move more rapidly. Whitehead wanted 
to get the Admiralties out of the way 
soon, so that he would have time to con- 
centrate against Wewak and Hollandia 
in the westward advance. Kenney, who 
had experience in New Guinea with 
quick seizures of airfields by light forces, 
had a scheme in mind for another such 
operation. Some time before 23 Febru- 
ary he told Whitehead to hit Los Negros 
hard but not to crater the runway. Hop- 
ing to force the Japanese to evacuate Los 
Negros and retire to Manus, he ordered 
frequent low-altitude photo-reconnais- 
sance missions. 8 

The Allies were not yet fully aware 
that Japanese air resistance in the South- 
east Area was almost a thing of the past, 
and that they had won. They knew, how- 
ever, that the enemy was weakening. The 
runways at Rabaul were usually cratered. 
On 21 February Allied intelligence rea- 
soned that Japanese aircraft were "ab- 
sconding" from Rabaul, probably to 
Truk and other bases in the Carolines. 9 
Further, five Seventh Fleet destroyers 

8 Kenney, General Kenney Reports, p. 358; Craven 
and Cate, The Pacific: Guadalcanal to Saipan, p. 

9 GHQ SWPA G-2 Daily Summary of Enemy Int, 
and GHQ SWPA G-z Est of Enemy Sit 700, 30-21 
Feb 44, both in GHQ SWPA G-3 Jnl, 21 Feb 44. 

sank a Japanese transport about one hun- 
dred miles east of Lorengau during a 
sweep on 22-23 February. Survivors 
testified that they were part of a 400-man 
detachment of air force ground crews 
that was being transferred to bases 
farther north. Three of the destroyers 
then sank a Japanese destroyer and a 
cargo ship south of New Hanover, 
skirted the southwest coast of New Ire- 
land, and steamed safely past Rabaul 
through Saint George's Channel, which 
lies between New Britain and New Ire- 
land, on the way back to base. The other 
two bombarded Kavieng. No Japanese 
aircraft opposed either group although 
these waters had formerly been domi- 
nated by Japanese air and surface forces. 

On 23 February— shortly after the 
great Truk raid and the withdrawal of 
Japanese naval aircraft from the South- 
east Area— Whitehead forwarded to Ken- 
ney a reconnaissance report from three 
B-25's that had just flown over Los 
Negros and Lorengau for ninety min- 
utes. Although they flew as low as twenty 
feet, they were not fired on, saw no Jap- 
anese, no trucks, and no laundry hung 
out to dry. The airfields were pitted and 
overgrown with grass. The whole area 
looked "completely washed out." White- 
head recommended that a ground recon- 
naissance party go in at once to check. 10 

When Kenney received this message 
he was at his headquarters in Brisbane. 
Concluding, with Whitehead, that "Los 
Negros was ripe for the plucking," he 
hurried to Mac Arthur's office and pro- 
posed to MacArthur, Kinkaid, and part 
of MacArthur's staff that a few hundred 
troops carried on APD's seize Los Negros 

10 Rad, Comdr AdVon Fifth AF to Comdr AAF, 
23 Feb 44, in GHQ SWPA G-3 Jnl, 23 Feb 44. 



and repair Momote airfield at once, 
rather than capture Seeadler Harbour, 
so that they could be reinforced and sup- 
plied by air if need be. This should be 
a reconnaissance in force. If resistance 
proved too strong the invaders could 
withdraw. A quick seizure of the Ad- 
miralties, Kenney argued, might make 
possible the bypassing of Kavieng and 
Hansa Bay. 11 

General Willoughby, in contrast with 
the airmen, was convinced that the Japa- 
nese garrison was fairly strong. His esti- 
mate for 25 February placed enemy 
strength at 4,050, 12 

MacArthur quickly decided in favor 
of the reconnaissance in force. Next day 
he radioed orders to Krueger, White- 
head, and Barbey to "prepare for im- 
mediate reconnaissance in force." He di- 
rected Krueger and Barbey to send eight 
hundred men of the 1st Cavalry Division 
and other units aboard two APD's and 
one destroyer division from Oro Bay to 
Momote not later than 29 February. If 
successful the cavalrymen were to pre- 
pare the airfield for transport aircraft 
and hold their positions pending arrival 
of reinforcements. 13 

The Reconnaissance in Force 


With but five days between Mac- 
Arthur's radiogram and D Day there was 
little time to make ready. But in accord- 

11 Kenney, General Kenney Reports, p. 359. 

12 Note, G-2 to G-3, 25 Feb 44, in GHQ SWPA 
G-3 Jnl, 25 Feb 44. GHQ SWPA G-3 Monthly 
Summary of Enemy Dispositions, 29 Feb 44, in GHQ 
SWPA G-3 Jnl, 29 Feb 44, gives the same figures. 

13 Rad, MacArthur to Comdr Alamo, CG AdVon 
Fifth AF, and Comdr VII Amphib Force, 24 Feb 44, 
in GHQ SWPA G-3 Jnl, 24 Feb 44. 

ance with GHQ's earlier orders, plan- 
ning had begun in January when 
Krueger directed the ist Cavalry Di- 
vision to prepare terrain, logistical, and 
intelligence studies. Krueger, White- 
head, and Barbey had begun a series of 
planning conferences on 19 February. 

Kinkaid's and Barbey's plans, issued 
on 26 February, provided for transport- 
ing and landing the reconnaissance force, 
then reinforcing it or withdrawing it if 
necessary. With the cruisers Nashville 
and Phoenix and four destroyers Rear 
Adm. Russell S. Berkey was to provide 
cover during the approach to the Ad- 
miralties and to deliver supporting gun- 
fire against Los Negros, Lorengau, and 
Seeadler Harbour during the landings. 
The attack group, which Barbey placed 
under command of Rear Adm. William 
M. Fechteler, his deputy, consisted of 
eight destroyers and three APD's. 1 * 

The cruisers were added to the force 
because General MacArthur elected to 
accompany the expedition, and to invite 
Admiral Kinkaid to go with him, to 
judge from firsthand observation whether 
to evacuate or hold after the reconnais- 
sance. The first plans had called for just 
one destroyer division and three APD's, 
but when Kinkaid learned of Mac- 
Arthur's decision he added two cruisers 
and four destroyers. This was necessary 
because a destroyer had neither accom- 
modations nor communications equip- 
ment suitable for a man of MacArthur' s 
age and rank. A cruiser would serve 
better, but a single cruiser could not go 
to the Admiralties. Kinkaid's policy for- 

14 Morison, Breaking the Bismarcks Barrier, pp. 
435-36; Rad, Comdr Seventh Fit to CTF's 76 and 
74, and CTG 74.2, 26 Feb 44, in GHQ SWPA G-3 
Jnl, 28 Feb 44. 



bade sending only one ship of any type 
on a tactical mission. Therefore he sent 
two cruisers, and the two cruisers re- 
quired four additional destroyers as es- 
corts. 15 

The air plans prescribed the usual mis- 
sions but necessarily compressed them 
into a few days. Bad weather limited the 
air effort on 26 February, but next day 
four B-25 squadrons attacked Momote 
and Lorengau while seven squadrons of 
B-24's attacked the Wewak fields and 
B-25's struck at Hansa Bay. Heavy at- 
tacks against the Admiralties and Hansa 
Bay followed on 28 February, and that 
night seven B-24's attacked Hollandia, 
far to the west. 16 

Krueger had originally planned to 
send a preinvasion reconnaissance party 
to the western tip of Manus, from where 
it was to patrol eastward for several 
weeks and radio reports to his head- 
quarters. But the new orders caused him 
to cancel this plan in favor of a recon- 
naissance on Los Negros. More data on 
Hyane Harbour and Japanese disposi- 
tions there and at Momote would have 
been useful, for there was still no agree- 
ment on enemy strength in that region. 
Willoughby's estimate of 4,050 conflicted 
sharply with that offered by Whitehead, 
who stated on 26 February that there 
were not more than 300 Japanese, mostly 
line of communications troops, on Los 
Negros and Manus. 17 And the 1st Cavalry 
Division's estimate placed Japanese 
strength at 4,900, despite an assertion 

" Oral statement of Adm Rinkaid to the author 
et al., 16 Nov 53. 

"Craven and Cate, The Pacific: Guadalcanal to 
Saipan, p. 563. 

11 Rad, Comdr AdVon Fifth AF to Comdr AAF, 
26 Feb 44, in GHQ SWPA G-3 Jnl, s>6 Feb 44. 

in its field order that "Recent air recon- 
naissance . . . results in no enemy ac- 
tion and no signs of enemy occupa- 
tion." 18 

Because Krueger did not wish to risk 
betraying Allied plans by sending a 
patrol to Hyane Harbour and Momote; 
he decided to send one to examine the 
region about one mile south of the har- 
bor. Accordingly, at 0645, 27 February, 
a PBY delivered a six -man party of 
Alamo Scouts to a point five hundred 
yards off Los Negros' southeast shore 
under cover of air bombardment. The 
scouts took a rubber boat ashore, found 
a large bivouac area on southeastern Los 
Negros, and reported by radio that the 
area between the coast and Momote was 
"lousy w r ith Japs." But when this report 
reached GHQ Kenney discounted it. He 
pointed out, with reason, that twenty-five 
enemy "in the woods at night" might 
give that impression and that the patrol 
had examined not the airdrome but only 
a part of the south end of Los Negros. 19 

This patrol did provide more data on 
which to base plans for naval gunfire 
support. Kinkaid and Barbey decided 
that one cruiser and two destroyers 
should fire on the bivouac area while the 
other cruiser and two more destroyers 
fired at Lorengau and Seeadler Harbour 
and other destroyers supported the land- 
ing itself. 

The 1st Cavalry Division, which was 

18 C£ par. ta (2) of Brewer TF FO 2, 25 Feb 44, 
with Annex i, Int, in Alamo Anchorage Jnl 3, 24- 
26 Feb 44. The reconnaissance forces estimated 
1,500 on Los Negros, altogether 4,350 in the Ad- 
miralties. Alamo FO 9 and Brewer TF FO 1 are 
orders prepared for the one-division invasion of 
the Admiralties scheduled for 1 April. 

rt Kenney, General Kenney Reports, p. 361. 



to provide the landing force, was unique 
in the U.S. Army in World War II. Dis- 
mounted and serving as infantry, it was 
a square division of two brigades plus 
division artillery. Each brigade consisted 
of two cavalry regiments of about two 
thousand men each. Each regiment was 
composed of headquarters, service, and 
weapons troops, and two squadrons. The 
squadrons contained a headquarters 
troop, three rifle troops, and a weapons 
troop. The weapons troop, using the or- 
ganization of the infantry heavy weapons 
company, had been added in 1943. 20 Di- 
vision artillery had a headquarters bat- 
tery, two 75-mm. pack howitzer battal- 
ions, and two 105-mm, howitzer battal- 

MacArthur's orders of 24 February 
specified that the landing force should 
number eight hundred men, including 
five hundred men of one squadron with 
additional artillery and service troops, 
but the next day he recommended that a 
slightly stronger force be used. 21 

On 26 February Krueger established 
the occupation force for the Admiralties 
as the Brewer Task Force. He placed it 
under Maj. Gen. Innis P. Swift, com- 
manding the 1st Cavalry Division, and 
assigned to it the ground force units pre- 
viously allotted by GHQ. 

For the D-Day landing Krueger and 
Swift organized the Brewer recon- 

20 The 1st Cavalry Division, a Regular Army unit 
with a high percentage of Regular officers and en- 
listed men, was not organized until igai, but its 
regiments had long and distinguished histories. The 
gth Cavalry (originally the ad Cavalry), organized 
in 1855, was commanded by Robert E. Lee, and 
the 7th Cavalry was George A. Custer's old 

51 Rad, MacArthur to Comdr Alamo, CG AdVon 
Fifth AF, and Comdr VII Amphib Force, 25 Feb 
44, in GHQ SWPA G-3 Jnl, 25 Feb 44. 

naissance force under Brig. Gen. William 
C. Chase, commander of the ist Cavalry 
Brigade. It consisted of detachments 
from 1st Cavalry Brigade Headquarters 
and Headquarters Troop; the 2d Squad- 
ron, 5th Cavalry; two 75-mm. howitzers 
of B Battery, 99th Field Artillery Bat- 
talion; the 673d Antiaircraft Artillery 
Battery (.50-caliber machine guns); the 
1st Platoon, B Troop, 1st Medical Squad- 
ron; the 30th Portable Surgical Hos- 
pital; air and naval liaison officers and 
a shore fire control party; and a detach- 
ment of the Australian New Guinea Ad- 
ministrative Unit— or about a thousand 

If the landing succeeded and the re- 
connaissance force stayed, the Brewer 
support force, under Col. Hugh T. Hoff- 
man, was to land on D plus 2. Hoffman's 
command embraced the remainder of 
the 5th Cavalry and the 99th Field Artil- 
lery Battalion, in addition to C Battery 
(90-mm.), 168th Antiaircraft Artillery 
Battalion; A Battery (multiple .50- 
caliber mount), 211th Antiaircraft Artil- 
lery Battalion; medical, engineer, and 
signal units from the 1st Cavalry Di- 
vision; and E Company, Shore Battalion, 
5 9 2d Engineer Boat and Shore Regi- 
ment. The 40th Naval Construction Bat- 
talion and detachments from other ele- 
ments of the 4th Construction Brigade, 
all from the South Pacific, were to ac- 
company the Support Force. 

The remainder of the Brewer Task 
Force, including the rest of the 1st 
Cavalry Brigade and the 2d Cavalry Bri- 
gade, was to follow if needed as soon as 
shipping became available. To shorten 
sailing time, Cape Cretin was to be used 
as a staging area for reinforcements. 

Hyane Harbour, scene of the initial 



Aboard the Cruiser Phoenix, 28 February 1944. Left, gunfire directed at 
Japanese heavy guns; right, Admiral Kinkaid and General MacArthur viewing 
the bombardment. 

landing, was indeed an unlikely place. 
Two small points of land about 750 
yards apart flanked the entrance; from 
them the enemy could put cross fire 
against landing craft sailing through the 
narrow gap in the barrier reef. Much of 
the shore line inside the harbor was 
covered by mangroves, but on the south, 
150 yards behind Momote airfield, a 
1,200-yard sandy beach with three jetties 
offered passage to troops and vehicles. 

(Map 20) 

With H Hour set for 0815 to give 
bombers time to deliver heavy strikes in 
support of the landing, the APD's were 
to anchor five thousand yards off Hyane 
Harbour. The destroyers carrying troops 
would enter the transport area to unload 
their passengers, then return to their fire 
support stations. Twelve LCP(R)'s were 

to carry the reconnaissance force ashore. 
The first three waves of four craft each 
would go in at five-minute intervals, un- 
load, return forty minutes later, and de- 
part again in three waves five minutes 
apart until the troops were ashore. 

To join the expedition, MacArthur 
and Kinkaid flew to Milne Bay and 
boarded the cruiser Phoenix in the after- 
noon of 27 February. The same afternoon 
at Oro Bay, where the 1st Cavalry Di- 
vision had been unloading ships and re- 
ceiving amphibious training, the Brewer 
reconnaissance force boarded Admiral 
Fechteler's ships, 1 70 men per APD and 
about 57 men per destroyer. The ships 
departed Oro Bay in late afternoon and 
early evening of 28 February, rendez- 
voused with Berkey's cruisers and de- 
stroyers early next morning just south of 



Cape Cretin, and followed eleven miles 
behind Berkey through Vitiaz Strait and 
the Bismarck Sea. No enemy ship or 
plane made an appearance. The sea was 
calm, the sky heavily overcast, as the 
ships neared Hyane Harbour. 

The Landings 

Fechteler ordered his ships to deploy 
at 0723, 29 February. Cruisers and de- 
stroyers took their support stations and 
commenced firing at 0740. APD's in the 
transport area lowered landing craft 
which proceeded toward their line of de- 
parture 3,700 yards from the beach. 

The heavy overcast and generally bad 
flying weather prevented all but a hand- 
ful of Allied B-24's from reaching the 
Admiralties before H Hour. P-38's, 
B-25's, and smoke-laying reconnaissance 
planes arrived later, but before they 
could attack the overcast closed in so 
tightly that they could do nothing. "The 
Fifth Air Force had made its chief con- 
tribution in pointing out the opportun- 
ity." 22 

The first sign of the Japanese came at 
H minus 20 minutes when the first wave 
of landing craft reached the line of de- 
parture. As it passed through the en- 
trance, enemy 20-mm. machine guns on 
either side opened fire while heavier 
guns directed their fire against the 
Phoenix and the destroyers. The cruiser 
and the destroyer Mahan promptly si- 
lenced a gun on Southeast Point, and 
other vessels silenced the machine guns. 
According to Admiral Kinkaid this per- 
formance so thoroughly converted Gen- 

" Craven and Cate, The Pacific: Guadalcanal to 
Saipan, p. 564. 

eral MacArthur into a naval gunfire en- 
thusiast that he became more royalist 
than the king, and thereafter Kinkaid 
frequently had to point out the limita- 
tions of naval gunfire to the general. 23 

Support plans called for naval gunfire 
to stop at 0755 (H minus 20 minutes) 
so that B-25's could bomb and strafe at 
low altitudes, but at 0755 no B-25's 
could be seen nor could any be reached 
by radio. The ships fired, therefore, until 
0810, and then fired star shells as a signal 
that strafers could attack in safety. Soon 
afterward three B-25's bombed the gun 
positions at the entrance to the harbor. 24 

Thus supported by air and naval bom- 
bardment, the leading wave of landing 
craft, carrying G Troop, 2d Squadron, 
5th Cavalry, met little fire as it passed 
through the entrance and turned left 
(south) toward the beach. It touched 
down at 08 1 7, whereupon an enemy ma- 
chine gun crew on the beach scrambled 
back for cover. The first man ashore, 2d 
Lt. Marvin J. Henshaw, led his platoon 
across the narrow beach to take a semi- 
circular position on the edge of a coco- 
nut plantation. There were no American 
casualties, but several Japanese were 
killed as they hastily made off in the 
direction of the airstrip. 

The Japanese resumed their positions 
at the harbor entrance when the naval 
shelling ceased and fired at the LCP(R)'s 
as they returned to the APD's. The 
Mahan steamed to within a mile of shore 
and fired 20-mm. and 40-mm. guns at the 
southern point. She could not put fire on 
the point opposite because the LCP(R)'s 

a Oral statement of Adm Kinkaid to the author 
et al., 16 Nov 53. 

" Craven and Cate, The Pacific: Guadalcanal to 
Saipan, p. 564. 



First Wave of Landing Craft Unloading men of G Troop, 2d Squadron, 5th 
Cavalry, 29 February 1944. 

were in the way. As the second wave 
started through the entrance so much 
enemy fire came from the skidway and 
from the northern point that it turned 
back. The destroyers Flusser and Dray- 
ton then put their fire on the north point 
while the Mahan pounded the southern. 
When the enemy fire ceased, the landing 
craft re-formed, went through the pas- 
sage, fired their machine guns at the 
skidway, and landed 150 men of the 
second wave at H plus 8 minutes. 

The second wave then passed through 
the first about a hundred yards inland. 
The third wave, which with the fourth 
received enemy fire on the way in, landed 
at H plus 30 minutes, pushed southwest, 
and established a line just short of the 
airstrip that included most of the eastern 

revetment area. So far, except for firing 
at the boats, the Japanese had not fought. 
At 0900 General Chase radioed Krueger 
that a line had been established three 
hundred yards inland, and "enemy situ- 
ation undetermined." 25 By 0950 the 
squadron, commanded by Lt. Col. 
William E. Lobit, had overrun Momote 
airfield. The troopers found it covered 
with weeds, littered with rusty fuselages, 
and pitted with water-filled bomb 

While the beachhead was relatively 
peaceful, the landing craft continued to 
receive fire on the way in and out. The 
destroyers continued intermittent bom- 

* Serial 7, 0900, 1st Cav Brigade Unit Jnl, ag Feb 
44, Vol. Ill of 1st Cav Brigade Hist Rpt Admiralty 
Islands Campaign. 



sd Lt. Marvin J. Henshaw receiv- 
ing the congratulations of General 
MacArthur, who awarded him the 
Distinguished Service Cross, 29 Feb- 
ruary 1944- 

bardment of the harbor entrance for 
about six hours. By the time the 
LCP(R)'s of the third wave had returned 
to the APD's, four of the total twelve had 
been damaged by the enemy gunfire. Be- 
cause the landing force probably could 
not be evacuated without the LCP(R)'s 
(although emergency plans called for an 
APD to penetrate the harbor and evacu- 
ate the 2d Squadron), the landing craft 
abandoned their schedule and entered 
the harbor only when the destroyers had 
forced the enemy to cease fire. A heavy 
rainstorm, which prevented the few 

Allied planes over the target from bomb- 
ing, soon reduced visibility so much that 
the Japanese fire became ineffective. 

The entire reconnaissance force was 
unloaded by 1250 (H plus 4 hours, 35 
minutes). Caliber .50 antiaircraft guns 
and the two 75-mm. pack howitzers had 
been manhandled ashore. Two cavalry- 
men had been killed, three wounded. 
Five Japanese were reported slain. Two 
sailors of the landing craft crews were 
dead, three wounded. 

"Remain Here" 

By afternoon Lobit's squadron had ad- 
vanced over the entire airfield including 
the western dispersal area, an advance of 
thirteen hundred yards on the longest 
axis, without encountering any more 
Japanese. Patrols moved across the island 
to Porlaka and north to the skidway 
without seeing an enemy. But it was 
clear that the Japanese had not evacu- 
ated. Other patrols advancing to the 
south had found signs of recent oc- 
cupancy, such as three kitchens and a 
warehouse full of rations, and a captured 
document indicated that some two hun- 
dred antiaircraft artillerymen were 
camped nearby. 

General MacArthur and Admiral Kin- 
kaid came ashore about 1600. Mac- 
Arthur awarded the Distinguished Serv- 
ice Cross to Lieutenant Henshaw, in- 
spected the lines, received reports, and 
made his decision. He directed General 
Chase "to remain here and hold the air- 
strip at any cost." 26 Having "ignored 
sniper fire . . . wet, cold, and dirty 

28 Quoted in 1st Cav Brigade Hist Rpt Admiralty 
Islands Campaign, I, 3. There are other versions of 
MacArthur's statement in existence. 


with mud up to the ears," he and Kin- 
kaid returned to the Phoenix, whence 
Mac Arthur radioed orders to send more 
troops, equipment, and supplies to the 
Admiralties at the earliest possible mo- 
ment. 27 The cruisers and most of the de- 
stroyers departed for New Guinea at 
1729, leaving behind the destroyers Bush 
and Stockton to support the cavalrymen. 

Chase and Lobit had obviously con- 
cluded that the larger estimate of enemy 
strength, rather than the airmen's, was 
the right one. If all the Japanese they 
estimated to be on Los Negros should 
counterattack, the one thousand men of 
the reconnaissance force would find it 
very difficult to hold both the airfield 
and the dispersal area. An inland defense 
line, about three thousand yards long ex- 
clusive of the shore, would have been re- 
quired to defend them. Because such a 
line could not safely be held by one 
thousand men, Chase decided to pull 
back east of the airstrip. He set up a line 
about fifteen hundred yards long which 
ran from the beach southward for about 
nine hundred yards, then swung sharply 
east to the sea. The troops did not occupy 
Jamandilai Point in their rear, but 
blocked its base that night and cleared 
it the next morning. The position se- 
lected on the edge of the strip provided 
a ready-made field of fire to the west. 

In late afternoon the troopers or- 
ganized their defenses. The beachhead 
was too small to permit the 75-mm. pack 
howitzers to cover the area immediately 
in front of the lines. Consequently the 

" Comment by Lt Col Julio Chiaramonte, at- 
tached to Ltr, Gen Chase to Gen Smith, Chief of 
Mil Hist, 6 Nov 53, no sub, OCMH; Rad, CINC- 
SWPA to CTF 76, and CG's Alamo and Fifth AF, 
ag Feb 44, in GHQ SWPA G-g Jul, 1 Mar 44. 


Digging a Foxhole Through Coral 
Rock, near the airstrip, 29 February 

field artillerymen were turned into rifle- 
men. The .50-caliber antiaircraft guns 
were set up on the front line. Outposts 
were established in the dispersal area on 
the other side of the airstrip. The 
soldiers found digging foxholes even 
more arduous than usual, for the soil 
was full of coral rock. The Americans' 
defenses suffered from two weaknesses: 
the impossibility of field artillery support 
immediately in front of the line, and the 
lack of barbed wire. 

To remedy the latter, Chase urgently 
requested Krueger to arrange for an air- 
drop of barbed wire and stakes, as well 
as mortar and small arms ammunition, 
on the north end of Momote drome as 
soon as possible. 28 

!s Item 25, 1st Cav Brigade Unit Jnl, 29 Feb 44. 



Careful preparations for defense were 
more than justified, because except for 
the air force ground crews the Japanese 
had not been evacuating either Los 
Negros or the island group. The larger 
intelligence estimates had been correct. 
And the Japanese, warned by American 
submarines that sent "frequently lengthy 
operational messages" from south of the 
Admiralties in late February, had been 
on the alert for an attack. 29 Most of the 
Admiralty garrison was stationed on Los 
Negros, with the ist Battalion, 229th 
Infantry responsible for the defense of 
the airfield and Hyane Harbour. One 
battalion defended Lorengau. The Japa- 
nese, expecting attack through Seeadler 
Harbour, had let the Americans slip in 
through the back door, but now they 
planned to take action. When General 
Imamura found out about the invasion, 
he ordered Colonel Ezaki to attack with 
his entire strength. 30 Ezaki did not im- 
mediately use the 2d Battalion, ist In- 
dependent Mixed Regiment, but left it 
at Salami Plantation north of the skid- 
way. He directed the ist Battalion, 
229th, commanded by a Captain Baba, 
to attack that night and "annihilate the 
enemy who have landed. This is not a 
delaying action. Be resolute to sacrifice 
your life for the Emperor and commit 
suicide in case capture is imminent." 31 

As dusk fell Japanese riflemen and the 
American outposts began a fire fight, 
whereupon the outposts were recalled. 
Shortly afterward small groups of enemy 

"Southeast Area Naval Operations, III, Japanese 
Monogr No. 50 (OCMH), 66. 

30 8th Area Army Operations, Japanese Monogr 
No. 110 (OCMH), p. 135. 

31 This order, a copy o£ which was captured, is 
quoted in part in [Frierson] The Admiralties, p. 

began moving up against the 2d Squad- 
ron's line in a series of un-co-ordinated 
attacks. Relying chiefly on grenades, the 
enemy groups attacked in darkness. Some 
managed to infiltrate through the line 
and cut nearly all the telephone wires. 

Baba's battalion delivered its heaviest 
attack against the southern part of the 
perimeter. Some Japanese, using life 
preservers, swam in behind the Ameri- 
can lines and landed. Another group 
broke through along the shore at the 
point of contact of the left (east) flank 
of E Troop and the right (south) flank of 
the field artillery unit, which was hold- 
ing the beach. The Americans defended 
by staying in their foxholes and firing 
at every visible target and at everything 
that moved. 

Two Japanese soldiers moved through 
the darkness and penetrated to the vi- 
cinity of General Chase's command post. 
Before they could do any damage Maj. 
Julio Chiaramonte, force S-2, killed one 
and wounded the other with a subma- 
chine gun. 

By daylight of 1 March most of the 
enemy attackers had withdrawn. During 
the morning the infiltrators who had 
hidden themselves were hunted down 
and killed. Seven Americans were dead, 
fifteen wounded, as compared with sixty- 
six enemy corpses within the perimeter. 

So far the reconnaissance force had 
held its own. But because the support 
force would not arrive until the next day 
patrols pushed westward and northward 
to determine just what Japanese opposi- 
tion was to be expected. After moving 
an average distance of four hundred 
yards they encountered the enemy in 
some strength. Clearly, another attack 
was probable. 



Chase's situation was improved during 
1 March by the arrival of more ammuni- 
tion. As the weather had cleared, three 
B-25's dropped supplies at 0830. Later 
in the day a B-17 of the 39th Troop 
Carrier Squadron made two supply runs, 
and four B-17's of the 375th Troop Car- 
rier Group each dropped three tons of 
blood plasma, ammunition, mines, and 
grenades. The reconnaissance force re- 
ceived no barbed wire. 32 

Captured documents had indicated 
the location of many enemy defensive 
positions, and the patrols that went out 
in the morning brought back more data. 
By now the Americans knew that Los 
Negros' south coast possessed prepared 
positions, and that the western dispersal 
area, Porlaka, and the coast of Hyane 
Harbour from the 2d Squadron's perim- 
eter north to the skidway were fortified. 
In consequence the two supporting de- 
stroyers and the 75-mm. pack howitzers 
bombarded these areas. Starting at 1600 
Fifth Air Force planes bombed the dis- 
persal area, and at 1715, when antiair- 
craft guns near the south end of the air- 
strip fired on the planes, the Bush and 
Stockton pulled to within a thousand 
yards of the shore and shelled them. 
Several 4th Air Army fighter planes from 
Wewak appeared but failed in their ef- 
fort to drive off the American planes. 
The air bombardment flushed a body of 
Japanese, estimated one hundred strong, 
from cover in the dispersal area. When 
these men rushed east across the strip in 
an effort to escape the bombs, most were 
cut down by the cavalrymen's fire. 

35 Craven and Cate, The Pacific: Guadalcanal to 
Saipan, p. 565, state that barbed wire was dropped, 
but several days later Chase was still protesting that 
he had not received any. 

Otherwise enemy ground forces re- 
mained quiescent during most of the 
afternoon except for a seventeen-man 
patrol of officers and sergeants, led by 
Captain Baba, which had apparently in- 
filtrated the lines on the previous night. 
Baba's patrol came through heavy under- 
brush to within thirty-five yards of Gen- 
eral Chase's command post. When the 
Americans sighted the patrol, Chase and 
his executive officer, Col. Earl F. Thomp- 
son, directed the movements of Major 
Chiaramonte and four enlisted men who 
moved out to the attack. After Chiara- 
monte's party had killed several Japa- 
nese, the others committed suicide with 
grenades and swords. 

The Japanese varied their pattern by 
striking at the perimeter at 1700 in- 
stead of after dark in an attack that was 
weaker than the one of the night before. 
Daylight simplified the defenders' task 
in repelling the attack, which ceased at 
2000. Thereafter throughout the night 
small groups harried and infiltrated the 
lines. About fifty Japanese used life belts 
to cross the harbor entrance and attack 
the position at the base of Jamandilai. 
In the course of the action the field artil- 
lerymen fired three hundred 75-mm. 
rounds at the enemy and also killed 47 
Japanese within the artillery positions 
with small arms fire. All together, 147 
Japanese were killed within the perim- 
eter during the two night battles. 

Actually, the Japanese, though possess- 
ing numerical superiority, had never 
used their full strength and had not 
seriously threatened Chase's force, which 
still held its lines intact on the morning 
of 2 March. Recklessness, coupled with 
the skill and tenacity of the cavalrymen, 
had cost the Japanese their best chance. 



To the Shores of Seeadler Harbour 

Seizure of Momote Airfield 

Meanwhile, at Oro Bay and Cape Cre- 
tin Colonel Hoffman's support force, 
numbering about 1,500 ground combat 
troops and 428 Seabees, had loaded 
aboard six LST's and an equal number 
of 2d Engineer Special Brigade LCM's 
that were towed by the LST's. These ves- 
sels, escorted by Australian and Ameri- 
can destroyers and two minesweepers 
under Capt. E. F. V. Dechaineux of the 
Australian Navy, made a quiet voyage 
and stood into Hyane Harbour shortly 
after 0900 on 2 March. 

The two minesweepers and one de- 
stroyer steamed to the north of Los 
Negros in an attempt to force the 1,500- 
yard-wide entrance to Seeadler Harbour. 
They encountered such heavy fire from 
Japanese coastal guns on the guardian 
islands of Hauwei and Ndrilo that 
they retired. Captain Dechaineux then 
brought three more destroyers, which 
fired at the Japanese while the first three 
ships again unsuccessfully attempted to 
force the passage. Minesweeping would 
obviously have to await ships with 
heavier guns than those of destroyers 
and minesweepers. 

The LST's and LCM's made their way 
through the entrance to Hyane Harbour 
and beached, whereupon Japanese mor- 
tars and machine guns north of the skid- 
way opened up. The landing craft re- 
plied with their machine guns, and at 
the same time B-25's attacked the Japa- 
nese positions. 

In the midst of the din the combat 
troops walked ashore. Then bulldozers 
left the LST's and began building ramps 
to get the other vehicles ashore. Unload- 

ing was finished by 1700, and the LST's 
departed; the LCM's remained in Hyane 

Before the LST's left for New Guinea 
Chase requested that the destroyers put 
fire in the northern point of land at 
Hyane Harbour. Four ships each fired 
fifty 5-inch rounds from close range, but 
when the LST's started out of the harbor 
they met machine gun fire. They replied 
and made the open sea in safety. One de- 
stroyer and the two minesweepers took 
the LST's to New Guinea while four 
destroyers stayed in the vicinity of the 
Admiralties to intercept any Japanese 
seaborne attacks. 

Since Momote airfield was not yet in 
American hands, Chase assigned part of 
the 40th Construction Battalion to a de- 
fensive sector on the right (north) flank 
of the beachhead. The Seabees, meeting 
some rifle fire while moving into posi- 
tion, used their ditch-digger to scoop out 
a 300-yard-long trench. 

As men, weapons, ammunition, sup- 
plies, and equipment came ashore during 
the day the beachhead became crowded. 
Chase decided to attack and enlarge his 
perimeter to include all the airfield, dis- 
persal area, and revetments, and the 
roads immediately around the airfield. 
When Hoffman came ashore in the 
morning he was met by Colonel Thomp- 
son, who took him to Chase's headquart- 
ers where the three officers completed 
plans for an attack by the 5th Cavalry 
that afternoon. 

At 1415, B-25's, P-38's, and P-47's 
began bombing and strafing the west half 
of the airfield, the dispersal area, the 
skidway, and the northern part of Los 
Negros. This attack lasted until 1530. 

The 1st and 2d Squadrons of the 5th 



Cavalry, on the left (south) and right 
respectively, had attacked at 1 500. There 
was no opposition from the enemy; 
within the hour Hoffman's regiment was 
in possession of the entire airfield and 
had begun to dig in along the line of 
the western and southern dispersal bays. 
The day's sole casualties, two men killed 
and four wounded, were caused by three 
American bombs that fell on positions 
held by E Troop and antiaircraft artil- 

It was clear to the Americans that the 
Japanese garrison had not yet made its 
maximum effort, for papers found in the 
advance over the airfield indicated that 
Baba's battalion was still south and west 
of the airfield. And earlier estimates had 
placed two thousand troops in the west 
half of Los Negros and Lorengau. Major 
Chiaramonte therefore warned the in- 
vasion force to expect attacks from the 
south, from Porlaka in the west, and 
southward from the skidway. 3S 

The invading Americans carefully pre- 
pared their defense positions. The front 
lines, still without barbed wire, included 
nearly all the dispersal area. Two anti- 
aircraft batteries and E Company, 593d 
Engineer Boat and Shore Regiment, 
were assigned to beach defense. Seabees 
established an inner defense line west 
and northwest of Chase's command post. 
The three 75-mm. batteries of the 99th 
Field Artillery Battalion set up in revet- 
ments some five hundred yards behind 
the front in a semicircle with overlapping 
sectors of fire. Because it was next to im- 
possible to prevent the Japanese from 

33 Brewer Ren TF S-2 Periodic Rpts 1-3, 1, 2, 
and 3 Mar 44, in Vol. II, 1st Cav Brigade Hist Rpt, 
27 Feb— 18 May 44. 

infiltrating the front lines, all units in- 
side the perimeter prepared all-round 
local defenses. 34 

Ezaki Attacks Again 

Colonel Ezaki now was preparing for 
a larger effort. He planned a co-ordinated 
attack, with the 2d Battalion, 1st Inde- 
pendent Mixed Regiment, driving south 
from Salami across the skidway, while 
one company, having moved from 
western Los Negros to Porlaka, struck 
eastward. Meanwhile, other detachments 
from the outlying islands and from in- 
land regions of Manus were to concen- 
trate at Lorengau. His forces were slow 
in concentrating, and Ezaki postponed 
the attack until the night of 3-4 March. 
As a result the 5th Cavalry was merely 
harassed in its new positions on the night 
of 2-3 March. 

The Americans used the daylight 
hours to strengthen their defenses. 35 
Bulldozers cleared fields of fire in front 
of the cavalry squadrons' lines. To keep 
infiltration to a minimum, each cavalry 
troop posted three rifle platoons in line 
with troop heavy weapons attached to 
each platoon. Japanese revetments were 
used as much as possible. Riflemen dug 
foxholes on the reverse slopes of the 
mounds, mines were laid in front, and 
the approaches to all positions were 
rigged with empty C-ration cans that 

34 General Chase again requested barbed wire by 
air but reported that he never received it. 

33 [Frierson] The Admiralties, pp. 43-44, asserts 
that when a group of Japanese officers attempted 
to land on the beachhead from a boat on the morn- 
ing of 3 March they were all killed, and that a 
document carried by one of them warned the 
Americans to expect attack that night. No contem- 
porary evidence to support this statement has been 



contained lumps of coral and were hung 
close to the ground so they would rattle 
when struck by a shoe. The 6o-mm. mor- 
tars were situated to deliver close sup- 
port fire directly in front of the cavalry 
squadrons, while the 8i's were massed 
near the center of the perimeter in front 
of the field artillery to deliver deeper 
supporting fire. Most of the antiaircraft 
.50-caliber machine guns were returned 
to their normal missions, but since the 
main attacks were expected from the 
north and west the guns posted on the 
north end of Momote field facing the 
skidway remained at the front. 

While the riflemen made ready, the 
artillery and the offshore destroyers fired 
at every evidence of the Japanese. They 
put concentrations on enemy groups 
north of the skidway. At 1600 field and 
antiaircraft artillery shot up several 
enemy barges that were observed behind 
overhanging vegetation on the north 
shore of Hyane Harbour. 

After dusk Japanese patrols began 
probing the lines, and at 2100 a lone 
plane dropped eight bombs which cut the 
telephone wires between the 1st Squad- 
ron and the 5th Cavalry command post. 

When the plane departed, flares and 
tracers heralded an attack by the rem- 
nants of the late Baba's battalion against 
the southwest portion of the perimeter, 
held by the 1st Squadron under Lt. Col. 
Charles E. Brady. Mortar and machine 
gun fire supported the attack, but it was 
weak. American mortars and machine 
guns beat off the attackers, although 
some infiltrated the lines, concealed 
themselves, and had to be flushed out 
and killed after daylight. 

The 2d Battalion, 1st Independent 

Mixed Regiment, delivered the main as- 
sault from the skidway, which was 
coupled with a drive east from Porlaka 
by other detachments. F Troop, which 
held the north-south portion of the line 
in the western dispersal area, and G 
Troop, defending the line from F 
Troop's right flank to the beach, received 
the brunt of the attacks. E Troop suf- 
fered only harassing attacks and infiltra- 
tion. By now the 2d Squadron, having 
landed on 29 February, had had more 
than enough experience in repelling 
night attacks, but this one differed from 
earlier ones in which the enemy had 
moved quietly and concealed himself as 
much as possible. On this night the Japa- 
nese advanced in the open in frontal as- 
sault with a good deal of talking, shout- 
ing, and even singing. Artillery and mor- 
tars opened fire at once. 

As they approached F and G Troops, 
the leading enemy waves hurled gre- 
nades, but they fell short of the cavalry 
lines. The Japanese pushed through the 
mine fields, taking casualties but not 
stopping, and drove into the interlock- 
ing bands of fire from the machine guns, 
which promptly cut them down. More 
kept coming; the cavalry lines held, but 
some Japanese managed to infiltrate and 
cut telephone lines. G Troop's three pla- 
toons stayed down in their positions and 
fired or hurled grenades at all possible 
targets. Just before dawn some Japanese 
soldiers penetrated G Troop's positions 
and Capt. Frank G. Mayfield organized 
a quick counterattack and drove them 
out. A few minutes later the Japanese 
assaulted again. This time, as two of 
Mayfield's platoons had exhausted their 
machine gun ammunition, the Japanese 



nearly succeeded in breaking through. 
But the Japanese were killed or driven 
off by a platoon of H Troop heavy ma- 
chine guns under S/Sgt. Edwin C. Terry. 

During these attacks Sgt. Troy A. Mc- 
Gill, of G Troop, was holding a revet- 
ment with his squad of eight men. When 
all but McGill and one other man had 
been killed or wounded, McGill ordered 
the other survivor to retire, fired his rifle 
until it jammed, then fought in front of 
his position with clubbed rifle until he 
was slain. McGill's gallantry won him the 
Medal of Honor. 38 

The attacks had been delivered with 
frequency and resolution throughout the 
night, but there was little evidence of 
skill or co-ordination. For example, 
about an hour before daylight a column 
of soldiers advanced down the road from 
Porlaka, singing, the cavalrymen later 
reported, "Deep in the Heart of 
Texas." 37 Mines, machine guns, rifles, 
and grenades killed nearly all of them. 

Reports of the night's action also re- 
late instances of Japanese shouting false 
commands in English and tapping tele- 
phone lines. One H Troop mortar sec- 
tion thought it heard an order to retreat 
and abandoned its position with the re- 
sult that the 2d Squadron lost its 81 -mm. 
mortar support. 

During the night the 99th Field Artil- 
lery Battalion fired almost continuously, 
as did all mortars except those that were 

"WD GO 35, 9 May 45. 

" The Japanese may have been singing, but it 
seems improbable that they sang this song, which 
was a favorite of the 1st Cavalry Division. The sd 
Squadron had been on Los Negros for four days, 
and this was its third night in close combat. One 
may guess that nerves were strained, imaginations 

abandoned. This fire was delivered in 
spite of harassing attacks from Japanese 
who had slipped through the front lines. 
Three field artillerymen were killed by 
infiltrators, and one antiaircraft crew 
abandoned its gun under pressure from 
the Japanese. Five Japanese, one with a 
grenade discharger, actually posted 
themselves on the roof of the dugout 
containing Colonel Lobit's command 
post, but Capt. Bruce Merritt killed 
them from his nearby foxhole. The Sea- 
bees, in their secondary line behind G 
Troop, passed ammunition to the hard- 
pressed cavalrymen and toward dawn 
some moved up to help G Troop hold 
its line. Other Seabees met a group of 
Japanese attacking two antiaircraft gun 
positions and killed them. 

By daylight of 4 March the Japanese 
had pulled back and the close fighting 
was over, but enemy mortars and field 
pieces hit the American positions until 
about 0730. The intensity of the night's 
action is indicated by the fact that two 
of the machine guns in G Troop's sector 
had fired a total of 8,770 rounds, and 168 
enemy corpses lay directly to the troop's 
front. There were no prisoners. Sixty- 
one Americans were killed, 244 
wounded, of whom 9 dead and 38 
wounded were Seabees. Ezaki had made 
his greatest offensive effort and failed. 
With more Americans due soon, the 
shattered Japanese units would be ca- 
pable of defensive action only. 

The Advance 

On 1 March, meanwhile, at Alamo 
headquarters on Cape Cretin General 
Krueger had completed plans for rein- 
forcement of Chase's men and for seiz- 



ure of the entire Admiralties group. 38 
Krueger ordered Swift to strengthen the 
reconnaissance force, seize Seeadler Har- 
bour, extend control over the entire Ad- 
miralties, and start building airdromes 
and a naval base. 39 On the 2d, the day 
Hoffman's support force landed, Krueger 
received an urgent request from Chase, 
who asked for his other regiment, the 
12th Cavalry. Krueger, Swift, and Bar- 
bey then arranged to speed up the move- 
ment to the Admiralties and land the 
12th Cavalry and other units on 6 
March, the 2d Cavalry Brigade on 9 
March instead of on g and 16 March, as 
they had originally planned. 40 They also 
arranged to rush the 2d Squadron and 
the Weapons Troop, 7 th Cavalry, and 
the 82d Field Artillery Battalion on 
three APD's, to arrive on the morning of 
4 March. 41 

General Krueger desired that Seeadler" 
Harbour be opened up. Two factors, 
besides the obvious one that Allied forces 
were eventually to use the harbor as a 
major naval base, motivated him. Hyane 

38 The total forces involved included the rest of 
the 1st Cavalry Division; Headquarters and Head- 
quarters Battery, 15th Antiaircraft Artillery Group; 
C Battery, 837th Antiaircraft Artillery Battalion 
(Searchlight); the 211th Coast Artillery Battalion 
(Antiaircraft); the 2d Battalion, 50th Coast Artil- 
lery Regiment; the Shore Battalion and A Com- 
pany, Boat Battalion, 59«d Engineer Boat and 
Shore Regiment; and a large array of signal, medi- 
cal, quartermaster, and engineer units. 

39 Alamo FO 11, 1 Mar 44, in Alamo Anchorage 
Jnl 7, 1—3 Mar 44. 

40 Rad, CG Alamo to GHQ SWPA, 1 Mar 44, in 
GHQ SWPA G-3 Jnl, 1 Mar 44. 

41 [Frierson] The Admiralties, p. 43; Rad, Chase 
to CG Alamo, 2 Mar 44, in Alamo Anchorage Jnl 
7; Rad, Alamo to GHQ SWPA, 2 Mar 44, and Rad, 
CTF 76 to Comdr Seventh Fit, 2 Mar 44, both in 
GHQ SWPA G-3 Jnl, 2 Mar 44. Barbey proposed 
using LST's for the 4 March reinforcements, but 
Krueger persuaded him to send them on APD's. 

Harbour and the initial beachhead were 
becoming too congested to receive the 
2d Brigade, but Salami Plantation on 
the west shore of the northwest penin- 
sula of Los Negros offered a good land- 
ing place. Clearing the harbor would 
also make possible a shore-to-shore move- 
ment from Los Negros against Manus. 
Therefore air and naval bombardments 
of enemy positions on the northwest tip 
of Los Negros, and on the guardian 
islands of Koruniat, Ndrilo, and Hau- 
wei were arranged. 42 Krueger, on 3 
March, ordered Swift to proceed to Los 
Negros at once, to survey the situation, 
and to take command ashore. 

On the morning of 4 March, shortly 
after Ezaki's attack subsided and after 
supporting Allied destroyers had shelled 
the skidway and the region to the north, 
the 2d Squadron, 7th Cavalry, and the 
82d Field Artillery Battalion (75-mm. 
pack howitzers) landed at Hyane Har- 
bour. Chase decided to wait for more 
troops before attacking, and put the 2d 
Squadron, 7th, under Lt. Col. Robert 
P. Kirk, in the line to replace the weary 
men of the 2d Squadron, 5 th Cavalry. 
Except for minor harassing attacks, in- 
filtrations, and a one-plane bombing at- 
tack, the night of 4-5 March was quiet. 

General Swift, accompanied by the 
1st Cavalry Division's chief of staff, Col. 
Charles A. Sheldon, and the intelligence 
and operations officers, reached Hyane 
Harbour aboard the destroyer Bush on 
the morning of 5 March. He assumed 
command of the troops ashore at 1 100, 
but since the Bush was busy executing 
fire support missions he stayed aboard 

"Alamo Force Rpt, Brewer Opns, p. 11; Rad, 
CG Alamo to GHQ SWPA, 4 Mar 44, in GHQ 
SWPA G-3 Jnl, 4 Mar 44. 



Lt. Gen. Walter Krueger (front seat) with Brig. Gen. William C. Chase and 
Maj. Gen. Innis P. Swift on an inspection tour, Los Negros Island. 

until 1600 so as not to interrupt the 

Swift directed Chase's reconnaissance 
force to clear the major part of Los 
Negros from Momote to the north and 
west, and to be prepared to extend over 
the entire island. He instructed the ad 
Cavalry Brigade, due to arrive on 9 
March, to land at Salami Plantation, to 
be prepared to move to a point on 
Manus west of Lorengau, and to attack 
eastward against that airfield and secure 
the eastern half of Manus.* 3 

To carry out the instructions for seiz- 
ing all of Los Negros, the 2d Squadron, 

43 Brewer TF FO g, 5 Mar 44, in Alamo Anchor- 
age Jnl 8, 4-6 Mar 44. 

7th Cavalry, was ordered to attack north 
across the skidway on the afternoon of 
the 5th. Accordingly the 2d Squadron, 
5th, began relieving Colonel Kirk's 
squadron in the perimeter late in the 
morning. At 1 1 20, while the relief was 
being effected, the Japanese began a 
series of harassing attacks, followed after 
noon by a resolute attack from Porlaka 
and the skidway. The enemy soldiers 
who broke into the front lines were all 
killed while field artillery and mortars 
broke up the attacks. Twenty-five dead 
Japanese bodies were counted, but 
twelve cavalrymen were wounded and 
it was 1630 before the 2d Squadron, 7th, 
was reorganized and ready to attack. 
Once it had moved beyond the perim- 



eter, Kirk's squadron found that the Jap- 
anese had mined the approaches to the 
skidway. The mines caused some casual- 
ties at first but thereafter were success- 
fully detected and removed. The squad- 
ron advanced slowly past enemy corpses 
that littered the road, but by darkness 
had reached the skidway, where it halted 
for the night. 

Kirk resumed his advance early on 6 
March. Later in the morning the 13th 
Cavalry Regimental Combat Team, 
commanded by Col. John H. Stadler, 
Jr., came ashore. Transported to Hyane 
Harbour aboard four LST's, the combat 
team, 2,837 men strong, consisted of the 
12th Cavalry; the 271st Field Artillery 
Battalion (105 -mm. howitzers); three 
light tanks of the 603d Tank Company; 
five LVT's of A Company, 5g2d Engi- 
neer Boat and Shore Regiment; and 
engineer, medical, and signal troops. 44 
When it reached shore, the 12 th Cav- 
alry, accompanied by the tanks, began 
moving north across the skidway to join 
Kirk in the advance, while the 271st 
Field Artillery Battalion moved into po- 
sition near the airstrip. The Japanese, 
obviously in retreat, offered only minor 
resistance. The advance was slowed 
chiefly by mud and trees they had felled 
across the roads and trails to Salami. 
Near the beach at Salami some Japanese 
in bunkers and buildings offered fight 
but were blasted out by tanks and 75- 
mm. howitzers. By 1630 all three cav- 
alry squadrons were established at Sa- 
lami. The surviving Japanese had es- 
caped by boat and canoe to the west. 
Thus by the day's end the 1st Cavalry 

14 12th Cav, Hist o£ lath Cav During the Ad- 
miralty Islands Campaign, 27 May 44, p. 3. 

Brigade held the beachhead where the 
ad Brigade was to land. 

Meanwhile air and naval surface 
forces had been at work on the Japanese 
positions guarding the entrance to 
Seeadler Harbour. Two days after the 
minesweepers were driven off, cruisers 
and destroyers of Admiral Crutchley's 
Task Force 74 bombarded suspected 
enemy gun positions on orders from 
Admiral Kinkaid, and on 5 March they 
fired eighty 8-inch, three hundred 6- 
inch, and one hundred 5-inch rounds 
without meeting any return fire. 45 Next 
morning, the lone destroyer Nicholson 
approached the harbor entrance to draw 
enemy fire. The Japanese opened up at 
850 yards range, whereupon Task Force 
74 and Allied bombers struck at the 
enemy positions thus disclosed. They 
were bombed again on 7 March by 
seven B-24's, and on 8 March by sev- 
enteen B-24's and eleven B-25's. There- 
after LCM's, destroyers, and other craft 
entered the harbor freely without en- 
countering enemy fire. 

From 6 through 8 March the 5th Cav- 
alry extended its holdings around the 
airstrip. The 2d Squadron took Porlaka 
on 6 March, then crossed Lemondrol 
Creek in canvas and rubber boats and 
amphibian tractors to seize Papitalai vil- 
lage on 7 March. 

American control over Seeadler Har- 
bour was furthered on 7 and 8 March 
by the seizure of two promontories 
northwest of Papitalai. The 2d Squad- 
ron, 12th Cavalry, using amphibian 
tractors, shuttled from Salami to Papi- 
talai Mission and captured it against 

"Rad, Comdr Seventh Fit to CTF 74, 4 Mar 44, 
and Rad, CTF 76 to CG Alamo, 5 Mar 44, both in 
Alamo Anchorage Jnl 8, 4-6 Mar 44. 



LST's Loaded With Troops and Equipment landing at Salami Plantation. 

sharp opposition. The gd Squadron, 7th, 
using LCM's, took Lombrum Planta- 

The 12 th Cavalry and the tanks pa- 
trolled to the northwest tip of Los Ne- 
gros to cover the 2d Brigade's landing, 
releasing, in the process, sixty-nine Sikh 
soldiers that the Japanese had been us- 
ing as laborers. 

On the morning of 9 March destroyers 
shelled Lorengau and minesweepers 
checked Seeadler Harbour. Then six 
LST's and one cargo ship entered the 
harbor to land Brig. Gen. Verne D. 
Mudge's 2d Cavalry Brigade and at- 
tached units at Salami Plantation. 46 

"This force consisted of: the 2d Cavalry Brigade 
(less the 2d Squadron and Weapons Troop, 7th 
Cavalry); the 61st Field Artillery Battalion (105-mm. 
howitzers); various divisional and nondivisional en- 
gineer, medical, quartermaster, and ordnance units; 
B Battery, 168th Antiaircraft Artillery Battalion 

Los Negros was now firmly in Allied 
hands. The next task facing the combat 
troops was the seizure of Lorengau. 


Plans and Preparations 

General Swift had assigned responsi- 
bility for capturing Manus to the 2d 
Cavalry Brigade. General Mudge accord- 
ingly had his plans ready the day after 
his brigade landed at Salami Planta- 
tion. 47 Not much was known about Jap- 

(Gun); B Battery, 211th Coast Artillery Battalion; 
F Company, 592d Engineer Boat and Shore Regi- 
ment; an Australian New Guinea Administrative 
Unit detachment; and a detachment of No. 62 
Works Wing, RAAF. 

" 2d Cav Brigade FO 2, 10 Mar 44, in Hist of Hq 
Troop, 2d Cav Brigade, Admiralty Islands Cam- 
paign, 9 Mar— 18 May 44 (actually the 2d Brigade's 
report); 8th Cav FO 2, 13 Mar 44, in 8th Cav Hist 
Rpt, Admiralty Campaign, 6 Mar— 20 May 44, Sec 11. 



anese strength on Manus, but recon- 
naissance had shown that Lorengau air- 
drome and Lorengau village east of it 
were fortified. As Lugos Mission was 
practically undefended, Mudge decided 
to land there— about 3,000 yards west of 
the airdrome. (Map 21) The beaches 
selected, Yellow 1 and Yellow 2, lay 
west and east of the Liei River. Yellow 
1, of coral sand, was 700 yards long, 14 
to 26 yards wide, with swamps imme- 
diately behind it. Yellow 2, 100 yards 
long, gave access to Number 3 Road 
which led along the coast to Lorengau. 

Mudge assigned the assault to the 8th 
Cavalry, commanded by Col. William 
J. Bradley. It was to land in column 
of squadrons, the 1st Squadron in the 
lead. Troop A was to land in L VT's on 
Yellow 2 east of the Liei, C Troop from 
LCV's on Yellow 1 to the west. The 
7th Cavalry, less the 2d Squadron, would 
follow the 8th ashore. The 2d Squad- 
ron, 7th, would constitute the brigade 
reserve. C Troop, 8th Engineer Squad- 
ron, was to land on Yellow 1, improve 
the beaches, and bridge the Liei to 
connect the beaches. Once ashore the 
8th Cavalry was to send the 1st Squad- 
ron east along Number 3 Road against 
Lorengau airfield while the other moved 
inland to Number 1 Road, and then 
moved east against Lorengau village to 
keep the Japanese from escaping inland 
to the jungled mountains where they 
would have a great defensive advan- 
tage. 48 

The cavalry generals arranged with 
naval officers and with Capt. George F. 
Frederick of the 12th Air Liaison Party 
for ample air and fleet support of the 
landing, in addition to support by field 
artillery. The islets north of Los Negros 
would provide positions from which field 
artillery could support the 2d Brigade's 
advance east by firing across its front at 
right angles to the axis of advance, as 
had been done in New Georgia during 
the advance on Munda airfield. There- 
fore plans were prepared for sending 
patrols to Hauwei, the Butjo Luo group, 
and Bear Point on Manus, just west of 
Loniu Passage, to determine enemy 
strength and look for artillery positions. 
D Day for Manus was first set for 13 

The island patrols, consisting of de- 
tachments from the 302d Cavalry Re- 
connaissance Troop plus artillery offi- 
cers, left Salami on 1 1 March. Bear 
Point, though not occupied by the en- 
emy, had so poor a beach that artillery 
could not be landed. Butjo Mokau, the 
most northern of the Butjo Luo group, 
offered good artillery positions and bore 
no signs of enemy occupation. In late 
afternoon F Troop, 7th Cavalry, occu- 
pied both islands of the group. 

The Hauwei patrol, a platoon strong, 
left Salami aboard an LCV and a PT 
boat and landed on the western part of 
Hauwei. 49 After the patrol had moved 
a short distance inland, machine gun, 
mortar, and rifle fire struck it from the 

"The landing force consisted of: the 2d Cavalry 
Brigade; C Troop, 8th Engineer Squadron; a de- 
tachment of the Shore Battalion, 592d Engineer 
Boat and Shore Regiment; detachments of the ist 
Medical Squadron and the ist Signal Troop; two 
medium and three light tanks. 

"Motor Torpedo Boat Squadrons 18 and 21 had 
arrived in Seeadler Harbour with their tender 
Oyster Bay. They served as a "sneak and peak" and 
general utility organization. Morison, Breaking the 
Bismarcks Barrier, p. 446. 



front and both sides. The patrol made 
a fighting withdrawal to the beach, sup- 
ported by fire from the PT boat and 
the LCV. But by the time the cavalry- 
men made the beach, the PT, whose 
skipper had been wounded, had re- 
turned to its tender. Five men boarded 
the LCV, but the remainder were still 
embroiled with the enemy. Mortar shells 
and machine gun bullets wounded most 
of the men aboard the LCV, which 
struck a submerged coral reef two hun- 

dred yards from shore and sank, leaving 
the survivors floating in the water. When 
about six Japanese started to set up a 
machine gun on the beach, the cav- 
alrymen still on shore shot them with 
submachine guns, then took to the water 
and joined the survivors from the LCV. 
After three hours in the water, the eight- 
een men, suffering from exposure to the 
sun and water, were picked up by a PT 
boat while a destroyer shelled Hauwei. 
An LCM later picked up one more man. 



Six men of the reconnaissance troop and 
two artillerymen had been killed, three 
were missing, and every survivor was 
wounded as well as burned. 50 

A larger force was obviously needed 
for Hauwei, and the landing on Manus 
would have to be delayed if the artillery 
was to get into position in time to sup- 
port the landing. Further, naval officers 
had already counseled delay in order to 
provide more time to clear the sea lanes 
to Lugos Mission. 

Accordingly the 2d Squadron (less F 
Troop), 7th Cavalry, was selected to 
seize Hauwei. Supporting its attack 
would be destroyers, rockets, 105-mm. 
fire from the 61st Field Artillery Bat- 
talion at Mokerang Plantation north of 
Salami, and P-40's of No. 77 Fighter 
Squadron, RAAF, which had reached 
Momote on g March. The squadron 
boarded LCM's at noon, 12 March, set 
out for Hauwei, and landed under cover 
of the supporting bombardment at 1400. 
The squadron later reported that "the 
covering fire was not accurate and most 
missiles fell short in the sea." 51 

E Troop landed on the west shore 
under small arms fire while G Troop, 
debarking on the south, met machine 
gun fire. The Japanese had rigged trip 
wires to activate mines, but the soldiers 
detected and avoided them. Kirk's squad- 
ron then drove inland against rifle fire 
and by 1500 held a north-south line 
across Hauwei about three hundred 
yards from the western tip and one thou- 
sand yards from the eastern end. By now 

M Maj. B. C. Wright, The ist Cavalry Division in 
World War II (Tokyo: Toppan Printing Company, 
1947), p. 27. 

u 7th Cav, Hist Rpt 7th Cav, 2 Mar-18 May 44, 
Pt. B, p. 5. 

the whole squadron was ashore, and H 
Troop's 81 -mm. mortars were ready to 
fire. E Troop continued its advance but 
G stayed in place. As contact broke 
between the two troops, Colonel Kirk 
pulled E Troop back and dug in for 
the night. 

General Mudge arrived at 1600 and 
after receiving Kirk's report ordered C 
Troop from Salami to Hauwei, and alert- 
ed one medium tank to move to Hauwei 
next day. C Troop arrived by LCM at 
1800 and took up a support position. 
During the night Japanese on Pityilu 
fired 20-mm. guns at the 2d Squadron 
but hit no one. The 61st Field Artillery 
Battalion put one thousand rounds of 
harassing fire on the enemy's section of 

Next morning, at 0900, the tank ar- 
rived and Kirk assigned his reconnais- 
sance platoon as close support. The at- 
tack began at 1000 with C, E, and G 
Troops abreast from left (north) to right. 
On the right a bunker, manned by eight 
Japanese with two 7.7-mm. machine 
guns, grenade discharges, and rifles, 
withstood four direct mortar bursts and 
four 75-mm. shells before it crumbled. 
In the center E Troop enveloped a short 
trench equipped with machine guns, 
grenade dischargers, and rifles. With 
these positions reduced the troops moved 
rapidly. By noon the 2d Squadron had 
covered the whole island. Eight Amer- 
icans had been killed, forty-six were 
wounded. Forty-three dead Japanese, all 
sailors, were counted. Captured booty 
included two 5-inch naval guns and a 
range-finder. One gun had been hit by 
the earlier bombardments; the other was 
in firing condition. 



That afternoon the 6ist Field Artil- 
lery Battalion unloaded its 105-mm. 
howitzers from LCM's and next day set 
them up on the southwest side while 
the 271st Field Artillery Battalion em- 
placed its 105-mm. howitzers on the west. 
The 99th Field Artillery Battalion, 
meanwhile, had emplaced twelve 75- 
mm. pack howitzers and six 37-mm. anti- 
tank guns on Butjo Mokau on 13 March. 

The Landing at Lugos 

With- the artillery now in position, 
embarkation of the 2d Cavalry Brigade 
aboard twelve LCM's, seven LCV's, and 
one LST began shortly after 0400 on 
15 March. The LST and the smaller 
craft proceeded separately to the ren- 
dezvous area off Lugos Mission and as- 
sembled about 0800. 

The three supporting artillery bat- 
talions had begun firing intermittently 
at Lorengau village at 2100 the previous 
evening, and at 0830 they shifted their 
fire to Lugos Mission. Four destroyers 
lying offshore fired at the shore line 
between the Tingau River, west of Lu- 
gos, and Lorengau until 0900. At 0900 
eighteen B-25's from Nadzab arrived 
overhead and from 0907 to 0925 put 
eighty-one 500-pound bombs and fired 
more than 44,000 machine gun bullets 
on the beaches. At 0925, when the bomb- 
ers cleared the area, three engineer 
rocket boats covered the first wave's 

The LST had previously disgorged 
seven LVT's, and six of them (the sev- 
enth was a rocket boat) bore A Troop 
toward Beach Yellow 2 while LCV's 
carried C Troop toward Yellow 1 . When 
the craft were close to shore, a machine 

gun east of the beaches opened fire. 
LVT's, and engineer support craft, and 
two PT boats replied and the gun fell 
silent. LVT's and LCV's landed their 
troops without casualties, and almost ex- 
actly on schedule. 52 

The soldiers of A Troop left the 
LVT's and drove through Lugos Mis- 
sion toward Number 3 Road. The few 
Japanese in the area, mostly sailors, did 
not offer determined resistance and were 
killed by A Troop and by later mop-up 
squads. C Troop, to the west, met no 
opposition as it advanced to a ridge some 
eight hundred yards inland where it es- 
tablished defenses to cover the landing 
of the 2d Brigade. 

Colonel Bradley had ordered the bulk 
of the 8th Cavalry to land at Yellow 2 
if it proved suitable for LCM's and 
LST's, and succeeding waves landed so 
quickly that Yellow 2 quickly became 
congested. The LST, which carried 
troops, weapons, and vehicles but no 
bulk cargo, was unloaded in forty-five 
minutes. It retracted to return to Salami 
for the 7th Cavalry. 53 This regiment, 
commanded by Col. Glenn S. Finley, 
landed from the LST and LCM's in the 
afternoon and took over defense of the 

Meanwhile the 8th Cavalry had begun 
its two-squadron advance against Loren- 
gau airdrome and Lorengau village over 
Roads 1 and 3. 

"General Swift, who observed the landing from 
the deck of a PT boat, is reported to have noted 
that timing was off by one and one-half minutes 
because a heavy sea slowed the landing craft. 
Wright, The ist Cavalry Division in World War 
11, p. 28. 

K 8th Cav Hist Rpt, Admiralty Campaign, Nar- 
rative of Events, p. 6. 



The Advance East: 

The Airfield and the Village 

During the day the 2d Squadron, 8th, 
under Maj. Haskett L. Connor, made its 
way southward along a native track 
toward Number 1 Road. Tractors bor- 
rowed from the artillery towed supplies 
and ammunition, Japanese riflemen har- 
assed the soldiers as they toiled slowly 
upward over a continuous succession of 
ridges. It was 1500 before F Troop, in 
the lead, reached Number 1 Road, where 
it ran into fire from three Japanese posi- 
tions which covered the track's junction 
with the toad. Enemy mortars to the 
south added their fire, and Connor de- 
cided to dig in for the night. At his re- 
quest the 61st Field Artillery Battalion 
silenced the enemy temporarily while 
the squadron established night defenses 
about six hundred yards from the road. 

The next morning, 16 March, Gen- 
eral Mudge and Colonel Bradley visited 
the squadron and observed its attack 
which, supported by one light and two 
medium tanks, overran the positions and 
enabled Connor's squadron to move east 
along Number 1 Road. The tanks had 
been hauled through the jungle with 
the aid of a D-7 bulldozer which cut 
down grades, cleared undergrowth, and 
towed the tanks when they stuck. One 
tank and the bulldozer remained at- 
tached to the 2d Squadron on its ad- 
vance along the road. By late afternoon 
it had reached a position on the road 
about a thousand yards west-northwest 
of Lorengau village and eight hundred 
yards south of the airstrip. 

The chief obstacle to the 2d Squad- 
ron's advance was terrain. The 1st 
Squadron (less C Troop) had had to fight 

its way along the coast on Number 3 
Road. After landing on the morning of 
15 March, the 1st, under Maj. Moyers 
S. Shore, had started east along the road 
behind A Troop. The road led through 
heavy rain forest interspersed with man- 
grove swamp on low ground. "The re- 
cent rains had softened the red clay until 
it assumed a glue-like consistence which 
made the footing difficult and slowed 
. . . leading elements." 54 

About one mile out of Lugos A Troop 
was halted by three pillboxes. 55 With the 
beach on one side and mangrove swamp 
on the other, there was no space for 
maneuver. Without orders from the 
troop commander one squad attempted 
an unsupported frontal assault which 
failed. Major Shore then alerted B 
Troop to pass through A and assault 
upon completion of an artillery prepara- 
tion. From Hauwei 105's of the 271st 
Field Artillery Battalion swept the en- 
emy area with shells that burst as close 
as a hundred yards from the cavalry- 
men. B Troop attacked but was quickly 

Shore then asked for a tank, more ar- 
tillery fire, and a strike by RAAF P-4o's 
(armed with 500-pound bombs), which 
had been on station since before H 
Hour, and arranged for 81 -mm. mortar 
support. "The combination of fire and 
bombs" turned the trick. They "plowed 
the position into a mass of craters," and 
B Troop advanced past "the blasted 

M Ibid., p. 6. 

M 1st Cavalry Division reports and journals use 
the word "bunker" for almost every enemy position 
encountered. Japanese positions along the road 
and at Lorengau appear to have been, according 
to World War II terminology, earth-and-log pill- 



remains of the pillboxes and scattered 
parts of their tenacious occupants. . . ." fl6 

The three pillboxes had apparently 
constituted the airstrip's western de- 
fenses, for when they crumbled the ist 
Squadron moved freely down the road. 
By 1700 it had advanced out of the 
jungle and held a ridge among the palms 
overlooking the southwest corner of the 
airstrip. Two of the squadron had been 
killed, eleven wounded, in the course of 
the action. Forty Japanese were reported 
killed. During the night of 15-16 March 
Shore's squadron, which C Troop re- 
joined at 1800 after its relief at the 
beachhead by the 7th Cavalry, received 
rifle fire from Japanese in a palm grove 
between the airfield and the sea. 

Next morning, 16 March, Shore de- 
cided to hold up his attack while an A 
Troop platoon went north of the strip 
to clear out the enemy riflemen and C 
Troop moved along the south edge of 
the airdrome to reconnoiter enemy po- 
sitions there. It was noon before the A 
Troop platoon accomplished its mission 
and the 1st Squadron could move. 

Meanwhile C Troop, after advancing 
200 yards over a series of rolling coconut- 
studded ridges which lay at right angles 
to the axis of advance, was halted by 
machine gun fire from a ridge about 150 
yards to its front. The troop commander, 
Capt. Winthrop B. Avery, emplaced the 
heavy machine guns and 81 -mm. mor- 
tars which had been attached from the 
Weapons Troop and attempted a co- 
ordinated attack. One platoon was to 
make a frontal assault while a second 
platoon worked around the south flank. 
The frontal attack failed, but the flank- 

K 8th Cav Hist Rpt, Admiralty Campaign, Nar- 
rative of Events, p. 7. 

ing platoon, commanded by S/Sgt. Ervin 
M. Gauthreaux, literally gained the top 
of the enemy positions, threw grenades 
into two pillboxes, and flushed several 
Japanese. 57 

But at this point things went wrong. 
With the enemy threat removed from 
his left (north) flank, and aware that C 
Troop was held up, Shore decided to 
leave C Troop in place to hold the 
enemy while the remainder of A Troop 
followed its other platoon through the 
palms and squadron Headquarters 
Troop, B Troop, and elements of D 
Troop drove down the airdrome on 
C's left. As B Troop advanced in the 
open it was struck by fire from the very 
positions that Gauthreaux' platoon was 
straddling, whereupon it halted, with- 
drew, and as it carried its casualties back 
to safety returned the enemy's fire. But 
the fire hit Gauthreaux's platoon, and 
Avery was forced to order him off the 
Japanese positions. 58 

After four hundred 105-mm. rounds 
had pounded the enemy position, C 
Troop attacked frontally while B Troop 
completed its retirement. But the Jap- 
anese still remained in their positions 
on the ridge and broke up C Troop's 
attack. By now all elements of the squad- 
ron had been committed and the Amer- 
icans had advanced to about the center 
of the airstrip. 

General Mudge arrived on the scene, 
inspected the squadron, reconnoitered 
the front, and decided to relieve the ist 
Squadron, 8th, with the 7th Cavalry. 
During the relief, which was effected 

51 Gauthreaux received the Silver Star and a com- 
mission as 2d lieutenant. 

58 Tanks had advanced along the north side of 
the strip but did not fire on the C Troop platoon. 



Men of the 8th Cavalry moving a 37-mm. antitank gun to a firing position near 
Lorengau Village, 18 March 1944. 

about 1600, the 7th Cavalry lost five 
men killed and fifteen wounded. 

With the previous day's experience 
as a guide, General Mudge and Colonel 
Finley planned a co-ordinated attack for 
17 March. The 7th Cavalry and the 2d 
Squadron, 8th, were to take the re- 
mainder of the airstrip and push on over 
the Lorengau River to the village. The 
1st Squadron, 7th, with squads from the 
8th Engineer Squadron attached, was to 
seize the eastern end of the airdrome 
while the 2d Squadron, 7th, moved south 
of the strip to make contact with the 2d 
Squadron, 8th, and advance to the river 
on Number 1 Road. 

During the night of 16-17 March de- 
stroyers and field artillery battalions 
shelled the Japanese positions, and in 

the early morning twenty-four 81 -mm. 
mortars, two light tanks, and two 37-mm. 
antitank guns put their fire on the pill- 
boxes. An 81 -mm. mortar of D Troop, 
8th, attached to the 7th Cavalry, de- 
molished one pillbox and its .50-caliber 
and .30-caliber machine guns and crew 
of fifteen men with a direct hit. When 
the mortars ceased fire automatic weap- 
ons opened up, and the 1st Squadron, 
7th, assaulted. "At 1033 when our troops 
came out of their fox-holes there were 
numerous cries of 'Garry Owen' as the 
1st Squadron went into its first action 
against the Japanese." 59 

There was little resistance, since the 

81 7th Cav, Hist Rpt 7th Cav, 3 Mar-18 May 44, 
Pt. B, p. ii. "Garry Owen" is the 7th's regimental 



supporting fires had "practically wiped 
out all enemy resistance except for nec- 
essary mopping up of a few bunkers still 
remaining intact." 80 The ist Squadron, 
under Maj. James A. Godwin, quickly 
seized the ridge that had held up Shore's 
squadron the day before, then encoun- 
tered another ridge position slightly to 
the east. After artillery and mortars had 
pounded it, cavalrymen moved in and 
occupied it. Flame throwers destroyed 
the pillboxes that remained in action. 

Meanwhile noon found the two in- 
land squadrons in contact with each 
other. By 1300 all three squadrons were 
in contact and had resumed the east- 
ward advance. Only a few scattered Jap- 
anese opposed the move from the air- 
strip to the river, but emplaced mines 
caused some casualties and slowed the 
advance, so that it was 1500 before the 
three squadrons pulled up on a ridge 
on the west bank that overlooked the 
village. It was too late in the day to at- 
tack Lorengau, which the Americans had 
reason to believe was strongly defended. 
The 7th Cavalry's reconnaissance pla- 
toon had immediately crossed the river 
over the sandbar at its mouth, met fire 
from Japanese positions west of Loren- 
gau, and withdrawn. Landing craft 
bringing in supplies received fire from 
the hills above the village. And, on the 
person of a Japanese officer who died 
defending the airdrome, the Americans 
had discovered maps of the defenses of 
Manus which showed that Lorengau and 
the road leading overland through the 
villages of Old Rossum and Rossum 
were fortified. 

w 7th Cav, Hist Rpt 7th Cav, 2 Mar-18 May 44, 
Pt. B, p. 11. 

Lorengau lies in a cup-shaped valley 
surrounded by 400-foot-high hills. Most 
of the Japanese defenses faced seaward, 
although positions also covered the roads 
leading east, west, and south. As the 
Lorengau River was about sixty feet 
wide and ten to twenty feet deep in 
most places, the 2d Brigade's best ap- 
proach route led over the alluvial sand- 
bar at the mouth. The enemy had plant- 
ed mines, controlled by a master switch 
in a pillbox on the hillside, on the 
stretch of beach between the sandbar 
and the hills. They had put foxholes and 
machine gun emplacements about a hun- 
dred yards inland from the shore, and 
had built about twelve pillboxes in the 
hills. The attackers would have to cross 
the river and the beach in full view of 
the enemy positions, but two factors fa- 
vored an assault: repeated bombings and 
shellings had uncovered several Japanese 
positions so that they were visible, and 
the ridge taken in the 17 March attack 
provided good observation over Loren- 

The 2d Squadron, 8th Cavalry, was 
designated to make the attack with mor- 
tar and artillery support. At 1000, 18 
March, the reconnaissance platoon led 
out in single file followed by E, F, and 
G Troops. The move was unexpectedly 
easy; only scattered machine gun fire 
was directed at the reconnaissance pla- 
toon, which quickly cleared the beach 
and the rifle pits. It cut the master cable 
leading to the mines. Later a dead Jap- 
anese was found in a small pillbox with 
the detonator switch clutched in his 

The rifle troops received fire and some 
casualties while crossing the river, but 
got over rapidly. On the east shore they 



Crossing the Lorengau River over the sandbar at the mouth of the river, 
1 8 March 1944. 

the 2d Squadron, 8th Cavalry, captured 
it at a cost of seven wounded. 

Fighting in the Admiralties was not 
yet over; it was 18 May before General 
Krueger officially terminated the opera- 
tion. Los Negros was not cleared of the 
enemy until the end of March, and it 
took two squadrons, several tanks, P-40 
strikes, and a good deal of artillery fire 
before the 2d Brigade cleared Number 
2 Road to Rossum on 25 March. But 
the capture of Lorengau airfield, follow- 
ing the seizure of Momote, placed in 
Allied hands the main strategic objec- 
tives of the operation. During the en- 
tire operation (including the seizure of 
more outlying islands in April) the 1st 
Cavalry Division lost 326 men killed, 
1,189 wounded, and 4 missing. It re- 
ported burying 3,280 and capturing 75 

deployed and prepared to attack. E 
Troop was to assault the enemy center 
in Lorengau with F Troop echeloned 
to the right rear; G Troop was to take 
the hills beside the river. Artillery and 
81-mm. mortars hit the enemy once 
more, and when their fire ceased 60- 
mm.'s and machine guns opened up, 
whereupon the cavalry troops assaulted 
the bunkers with grenade, submachine 
gun, rifle, and flame thrower. Again, it 
was unexpectedly easy, for the Japanese 
apparently retreated inland over Num- 
ber 2 Road. 61 Eighty-seven Japanese 
were killed defending Lorengau, while 

01 Little is known about Japanese decisions and 
movements, as no Japanese survived to report to 
Rabaul. 8th Area Army Operations, Japanese 
Monogr No. no (OCMH), p. 134. 



Troop G, 8th Cavalry, Near Number i Road on the west side of the Lorengau 
River, 18 March 1944. 

of the enemy, and General Krueger es- 
timated that the Japanese had disposed 
of 1,100 more bodies. 62 

Base Development 

Meanwhile several battalions of Sea- 
bees, plus Army engineer units, were 
building airfields and a naval base. Mac- 
Arthur, Nimitz, and the Joint Chiefs 
of Staff had intended that the naval base 
be used by all Allied fleets serving in 
the Pacific. In February Nimitz proposed 
to Admiral King that Admiral Halsey, 
who furnished most of the service troops, 
be given responsibility, under Nimitz, 
for developing and controlling the 
base. 63 Nimitz' proposal was rejected by 

*° Alamo Force Rpt, Brewer Opns, p. 26. 
,a Rad, CINCPAC to COMINCH, 23 Feb 44, CM- 
IN 16947. 

the Joint Chiefs but not before MacAr- 
thur became so irate that he ordered 
work on the Admiralties "restricted to 
facilities for ships under his direct com- 
mand—the Seventh Fleet and British 
units." 64 Halsey, whom MacArthur 
vainly requested as his commander of 
Allied Naval Forces, made a hurried 
trip to Brisbane in early March and 
found that MacArthur "lumped me, 
Nimitz, King, and the whole Navy in a 
vicious conspiracy to pare away his au- 
thority." 05 Halsey was in a difficult posi- 
tion. MacArthur was very angry; he was 
Halsey's superior, and was vastly senior 

m Halsey and Bryan, Admiral Halsey's Story, p. 
189; Rad, Marshall to MacArthur, 9 Mar 44, CM- 
OUT 3710. 

05 Halsey and Bryan, Admiral Halsey's Story, p. 
189; Suppl Min, JCS mtg, 8 Feb 44; Rad, Mac- 
Arthur to Marshall, 2 Feb 44, CM— IN 1443. 



to him. 66 And it is probably gratuitous 
to say MacArthur was formidable in ar- 
gument. The scene, as Halsey records it, 
was lively, with MacArthur expressing 
himself strongly. "Unlike myself," the 
Admiral wrote, "strong emotion did not 
make him profane. He did not need to 
be; profanity would have merely discol- 
ored his eloquence." But the ram-jawed 
Halsey could also be formidable. Sup- 
ported by Kinkaid and Carney, he asked 
the General to rescind his order: 
". . . 'if you stick to this order of yours, 
you'll be hampering the war effort!'" 
Halsey went on to say that "the com- 
mand of Manus didn't matter a whit to 
me. What did matter was the quick con- 
struction of the base. Kenney or an Aus- 
tralian or an enlisted cavalryman could 
boss it for all I cared, as long as it was 
ready to handle the fleet when we moved 
up New Guinea and on toward the 
Philippines." After long argument, Gen- 
eral MacArthur agreed to cancel his or- 
der and the work went forward under 
Admiral Kinkaid's direction. 67 

"° During World War I MacArthur had com- 
manded a brigade and then a division while Halsey 
commanded a destroyer. 

61 Halsey and Bryan, Admiral Halsey's Story, pp. 
189-90; Oral statement of Adm Kinkaid to the 

Momote airfield, first used in March, 
was seven thousand feet long by 1 8 May. 
When the Lorengau airstrip proved un- 
suitable, Seabees and the 836th Engineer 
Aviation Battalion, working under pres- 
sure, finished one at Mokerang Planta- 
tion by 21 April, then put in a parallel 
runway. Seabees installed two runways 
for carrier aircraft on the outlying 
islands, and also developed Seeadler Har- 
bour into one of the largest naval bases 
in the Pacific, with repair facilities for 
all types of warships and transports. 68 As 
planned, the naval base serviced the 
Third, Fifth, and Seventh Fleets in later 
operations, and the airfields supported 
the drives along the New Guinea coast 
and through the Central Pacific. The 
gallant action of the 1st Cavalry Division 
in execution of MacArthur's bold de- 
cision thus paid rich dividends. 

author et al., 16 Nov 53. General Kenney, in his 
comments, remarked that Halsey's statement sounds 
as though "he didn't like me, the Australians, or 
enlisted cavalrymen." Actually Halsey was probably 
only listing people unlikely to be directing con- 
struction of a naval base. 

88 Building the Navy's Bases in World War II, 11, 
295-302; Off of Chief Engr, GHQ AFP AC, Airfield 
and Base Development, pp. 208-22, and Critique, 
PP. 145-53- 


Bougainville Counterattack 

By March 1944 the Japanese were 
clearly beaten in the Southeast Area. 
With air and naval strength gone, the 
ground troops were stranded, immobi- 
lized, incapable of affecting the course of 
the war. Only at Rabaul were the Japa- 
nese strong, and that strength could not 
be employed unless the Allies chose to 
attack. But among the characteristics 
that made the Japanese a formidable op- 
ponent was his refusal to accept defeat 
even in a hopeless situation. If beaten, 
he knew it not. Thus it was that Gen- 
erals Imamura and Hyakutake designed 
the destruction, in March, of the XIV 
Corps at Empress Augusta Bay, Bougain- 
ville. 1 

'This chapter is based on SOPACBACOM, The 
Bougainville Campaign, Chs. IV-IX, supplemented 
by rpts, jnls, and jnl files of XIV Corps, Americal 
Div, 37th Div, and the principal component units 
which participated; Maj Gen Oscar W. Griswold, 
Bougainville: An Experience in Jungle Warfare 
(typescript); ACofS G-a XIV Corps, History of the 
"TA" Operation, Bougainville, March 1944 [zi 
Apr 44]! 8th Area Army Operations, Japanese 
Monogr No. 110 (OCMH), pp. 106-22; 17th Army 
Operations, II, Japanese Monogr No. 40 (OCMH), 
105-29; Capt. Francis D. Cronin, Under the South- 
ern Cross: The Saga of the Americal Division 
(Washington; Combat Forces Press, 1951), pp. 143- 
68; Frankel, The 37th Infantry Division in World 
War II, pp. 141-70; Answers (27 Jul 49) of Gen 
Kanda [former CG, 6th Div~\ to questions by Hist 
Sec G-2 FEC, in Hist Div MIS GHQ FEC, State- 
ments of Japanese Officials on World War II (Eng- 
lish Translations), II, 19—31, OCMH, 


The Approach 

When in late 1943 the Japanese com- 
manders had finally concluded that the 
invasion of Empress Augusta Bay was 
actually the Allied main effort at Bou- 
gainville, they began making plans for 
their counterattack. Unfortunately for 
him, Hyakutake's intelligence estimate 
was as inaccurate as most other Japanese 
estimates during World War II. He 
placed Allied strength at Empress 
Augusta Bay at about 30,000 of whom 
10,000 were supposed to be aircraft 
ground crews. His figure for General 
Griswold's total strength was too low by 
half. Against the XIV Corps he planned 
to use the main strength of the ijth 
Army, which consisted principally of 
General Kanda's 6th Division and several 
battalions of the ijth Division that Ima- 
mura had sent down in November. Total 
Japanese strength involved is variously 
reported as 15,000 to 19,000 men. 2 

! ACofS G-2 XIV Corps, History of the "TA" 
Operation, a careful, conservative study written after 
the counteroffensive from prisoner-of-war interroga- 
tions, captured documents, and G— 2 periodic re- 
ports and summaries, gives 15,400 men as the total. 
In 1949 General Kanda, speaking from memory, 
said there were 19,000 men involved plus about 
2,000 sailors. He may have included all troops in 
rear areas in his figure. 



During the early part of 1944 Japa- 
nese engineers built or improved roads, 
trails, and bridges so that the ijth Army 
could move from north and south Bou- 
gainville to assembly areas in the hills 
inland from the XIV Corps' perimeter. 
By mid-February the enemy soldiers 
were all on their way, and Hyakutake 
left Erventa to supervise the action him- 

The Japanese had hoped to launch an 
amphibious assault against the Ameri- 
cans, coupled with an attack from inland. 
A shortage of landing craft made the 
amphibious assault impossible, but 
barges, operating on moonless nights to 
avoid Allied aircraft and PT boats, trans- 
ported heavy equipment, including artil- 
lery, to a point east of Cape Torokina 
from where it was laboriously hauled in- 
land to the hills. Packhorses and trucks 
carried supplies part of the way on the 
overland routes. 

The infantry regiments of the 6th Di- 
vision advanced along both coasts, the 
13th and 23d Infantry Regiments on the 
west, the 45th Infantry up the east coast 
to Numa Numa, thence southwest by the 
Numa Numa Trail. The ijth Division 
battalions also marched along both coasts 
from their positions in the north. 

Such a move could hardly go unno- 
ticed. Coastwatchers, radio intercepts, 
long- and short-range ground patrols, in- 
terrogation of prisoners and even of a 
few deserters, Japanese activity near the 
Fiji outpost at Ibu, interpretation of 
aerial photographs, and air and naval 
searches told General Griswold that the 
Japanese were on the move all over the 
island, and that attack was imminent. 
Allied planes regularly bombed all sus- 

pected troop movements, bridges, and 
assembly areas. When the Japanese 
launched strong attacks at Ibu in mid- 
February, the corps commander ordered 
the Fijians back to the perimeter. Four 
hundred and fifty Fiji soldiers and two 
hundred Bougainville natives made their 
way to Cape Torokina. Two Fijians 
were slightly wounded during the with- 

Patrol clashes and fire fights in the 
hills north and northeast of the XIV 
Corps' perimeter indicated that the Jap- 
anese were concentrating there. Further, 
Japanese carelessness in safeguarding im- 
portant documents played into General 
Griswold's hands. Papers taken from 
enemy corpses gave him a precise idea 
of Hyaku take's plan of attack, told him 
exactly which Japanese units were about 
to attack him, and gave him the general 
location of the enemy artillery units. In- 
formation about the attack was posted 
on the American units' bulletin boards. 

XIV Corps' Defenses 

At the beginning of March the XIV 
Corps' perimeter was somewhat larger 
than it had been when Griswold took 
over. It included, in a horseshoe-shaped 
line on the inland side, some 23,000 
yards of low hills and jungle. The beach 
frontage totaled 11,000 yards. Depth of 
the position was about 8,000 yards. (Map 
22) The main ground combat elements 
of the corps were the Americal and 37th 
Divisions, which numbered about 27,000 
men. All together, 62,000 men, including 
naval units, were attached or assigned to 
the XIV Corps. 

All the infantry regiments were placed 
on the front lines. A total of twelve rifle 



battalions held frontages varying from 
2,000 to 2,400 yards. Usually each regi- 
ment held one battalion in reserve. The 
37th Division defended the left (north- 
west) sector from a point on the beach 
5,500 yards northwest of Cape Torokina 
to the area of Hill 700, about 2,000 yards 
east of Lake Kathleen. The 148th In- 
fantry, on the division left, and the 129th 
Infantry, in the center, held low ground. 
The 145th Infantry, on the right, held 
Hill 700, the highest ground possessed 
by the Americans. The Americal Di- 
vision's line ran from just east of Hill 
700, where the 164th Infantry's left flank 
tied in with the i45th's right, over Hills 
608, 309, and 270, then along the west 
bank of the Torokina River. Near its 
mouth the line crossed over to the east 

bank. 3 The i82d Infantry, in the di- 
vision's center, held Hills 309 and 270 
on the main perimeter line. The i32d 
Infantry on the right held low ground. 
In addition a detachment of the i82d In- 
fantry, plus artillery and mortar ob- 
servers, maintained an outpost on Hill 
260, an eminence which was some dis- 
tance east of the main line of resistance 
and overlooked the Torokina River. 
Griswold had ordered this hill held so 
that it could be used as an American 
artillery and mortar observation post, 
and so that the enemy could not use it 
to observe American positions. 

'The 133d Infantry, Americal Division, had seized 
this area in an action in January wherein S/Sgt. 
Jessie R. Drowley fought so valiantly that he was 
awarded the Medal of Honor. WD GO 73, 6 Sep 44. 



All units had been developing and 
strengthening positions on the main line 
of resistance, which now consisted of rifle 
pits and earth, log, and sandbag pill- 
boxes, wired in behind double-apron or 
concertina barbed wire. In front of the 
wire were minefields. Various devices 
were employed to give illumination at 
night: searchlights, either shining di- 
rectly or reflecting a spread beam off 
clouds; flares tied in trees and set off by 
pull wires; flashlights; thermite gre- 
nades; and cans full of sand and gasoline. 
Grenades, with wires attached, were set 
up as booby traps along obvious ap- 
proach routes. Oil drums, each with 
scrap metal packed around a bangalore 
torpedo, were wired for electrical detona- 
tion. Fields of fire fifty yards or more 
deep, deep enough to prevent the enemy 
from throwing hand grenades at the 
American positions from cover and con- 
cealment, had been cleared. Almost all 
the infantry regiments possessed extra 
machine guns, and had issued two BAR's 
to each rifle squad. All regiments had 
constructed reserve positions. The naval 
construction battalions, the 3d Marine 
Defense Battalion, Army engineer units, 
and others maintained provisional in- 
fantry units as part of the corps reserve, 
which also included the 82d Chemical 
Battalion, the 754th Tank Battalion, and 
the 1st Battalion, 24th Infantry. 4 

Artillery support for the perimeter, 
though below American standards, was 
stronger than the enemy's supporting 
artillery. The XIV Corps still had 

4 This unit had served on Bougainville since 30 
January, chiefly as a labor battalion. See Ulysses G. 
Lee, Employment of Negro Troops, a volume in 
preparation for the series UNITED STATES 

neither organic artillery nor an artillery 
commander. Serving as corps artillery 
commander was General Kreber, artil- 
lery commander of the 37th Division. 
Under General Kreber's command were 
the eight (six 105-mm. and two 155-mm.) 
howitzer battalions organic to the two 
divisions, plus the provisional corps 
artillery. This consisted of two 155-mm. 
gun batteries of the 3d Marine Defense 
Battalion; four 90-mm. antiaircraft bat- 
teries of the 251st Antiaircraft Artillery 
Regiment; and four 90-mm. antiaircraft 
batteries of the 3d Marine Defense Bat- 
talion, of which one, D Battery, 70th 
Coast Artillery (Antiaircraft) Battalion, 
was attached from the Army, Gun power 
of the XIV Corps units was augmented 
on 3 March when six cannon companies, 
with 75-mm. pack howitzers, reached 
Bougainville and joined the infantry 

The XIV Corps' positions were 
strong, and since he possessed interior 
lines General Griswold could easily 
switch his reserve units back and forth. 
But the positions were not ideal. The 
corps lacked enough men, by American 
standards, to hold all the high ground 
in the vicinity. Beyond the coastal plain 
the ground rises abruptly from ridge to 
ridge, each higher than the preceding 
one, up to the summits of the Crown 
Prince Range. Thus the Americans on 
Hills 608 and 700 held positions that 
were dominated by the higher ground 
in Japanese hands— Blue Ridge, three 
thousand yards north of Hill 700, and 
Hills 1000 and 1111, just southeast of 
Blue Ridge. These hills gave the enemy 
an excellent view over all the perimeter 
except the reverse slopes of the Ameri- 
can-held hills. By 1 March, however, 



General Griswold was sure that "the 
perimeter was as well organized as the 
personnel and the terrain would per- 
mit." 5 

The Japanese Plan of Attack 

General Hyakutake organized most of 
his infantry into three forces, each 
named for its commander. The lwasa 
Unit, under General lwasa, consisted of 
the 2d Battalion, 13th Infantry; the 23d 
Infantry; and two batteries of field artil- 
lery, some mortars, and engineers and 
other supporting troops. The Magata 
Unit, led by Col. Isaoshi Magata of the 
45th Infantry, whom Kanda considered 
to be a crack regimental commander, in- 
cluded nearly all the 45th Infantry plus 
mortars, field artillery, and engineers. 
The third unit, under Col. Toyoharei 
Muda, who had succeeded the late 
Tomonari in command of the 13th In- 
fantry, consisted of the remainder of the 
13th plus engineers. Supporting the at- 
tacks of the three infantry units was an 
artillery group commanded by a Colo- 
nel Saito. This consisted of four 150 mm. 
howitzers, two 105-mm. howitzers, and a 
number of smaller pieces. 6 Artillery am- 
munition supply totaled three hundred 
rounds per piece, of which one fifth was 
to be used for direct support of the in- 
fantry, the rest for interdicting the air- 

Also present were elements of the 1st 
and 3d Battalions, 53d Infantry, and 
part of the 81st Infantry, all from the 
17th Division. At first these were either 

* Griswold, Bougainville, p. 46. 

'General Kanda specified 18 70-mm. battalion 
guns; ACofS G-2 XIV Corps, History of the "TA" 
Operation lists 168 75-mm. mountain guns. The 
Japanese 105 is often called a 10-cm. piece. 

placed in rjth Army reserve or were as- 
signed diversionary missions against the 
northwest part of the XIV Corps' beach- 

The hvasa Unit assembled behind 
Hill 1111, the Magata Unit behind 
Mount Nampei, a shoulder-shaped ridge 
extending outward from the Crown 
Prince Range just northwest of Blue 
Ridge. The Muda Unit assembled at 
Peko, a village on the East-West Trail 
about 5,400 yards east-northeast of Hill 
260. The artillery group emplaced in the 
vicinity of Hill 600. 

The plan of maneuver involved two 
thrusts from the north coupled w T ith an 
attack from the northeast, all on a com- 
plicated schedule. Briefly, the lwasa Unit 
was to attack and secure Hill 700 on Y 
Day (set, after some delays in moving 
into position, for 8 March), reorganize 
on 9 and 10 March, and advance to the 
Piva airfields. During this period the 
Muda Unit was to capture Hills 260 and 
309, whereupon it and one battalion of 
the lwasa Unit were supposed to attack 
Hill 608 from the southeast and north- 
west on 12 March. All these attacks were 
preliminary to an effort which was to 
be delivered, starting 1 1 March, by the 
Magata Unit against the 129th Infantry 
in its low ground west of Hill 700. 
Magata's men, after cracking the i2gth's 
line, were to advance against the Piva 
airstrips in conjunction with Iwasa's ad- 
vance. Then all units were to drive 
southward on a broad front to capture 
the Torokina fighter strip by 17 March. 
Haste was essential, since the ijth Army 
had brought with it but two weeks' 

Hindsight indicates that the Japanese 
plan was unsound. Even had Hyaku- 



155-MM. Guns of the 30 Marine Defense Battalion firing on enemy positions, 
6 March 1944. 

take's estimate of American strength 
been correct, he still lacked enough 
strength in manpower and in artillery 
(he had no air support whatever) to at- 
tack prepared positions, and under the 
actual circumstances he was hopelessly 
outnumbered and outgunned. If his ob- 
ject had been to inflict maximum dam- 
age regardless of his own losses, he might 
have achieved a larger degree of success 
by concentrating his forces from the first 
in order to overwhelm a narrow portion 
of Griswold's front, break through, and 
spread destruction throughout the rear 
areas until Griswold could redeploy his 
infantry regiments. Of course, Hyaku- 
take might have achieved more success 
had the American soldiers elected to turn 

and run instead of standing their 
ground, but that was an imponderable 
that he could not count on. The Amer- 
ical and 37th Divisions were veteran 
units. 7 

By 8 March almost everything was 
ready. The rhetorical manifestoes by 
which Japanese officers exhorted their 
troops were issued. General Hyakutake 
expressed himself along these lines: 

The time has come to manifest our 
knighthood with the pure brilliance of the 
sword. It is our duty to erase the mortifica- 

' Hyakutake, in 1942, had delivered a similar, un- 
successful, counterattack against Vandegrift's posi- 
tions on Guadalcanal. See Miller, Guadalcanal: The 
First Offensive, Ch. VI. 



tion of our brothers at Guadalcanal, Attack! 
Assault! Destroy everything! Cut, slash, and 
mow them down. May the color of the red 
emblem of our arms be deepened with the 
blood of the American rascals. Our cry of 
victory at Torokina Bay will be shouted re- 
soundingly to our native land. 

We are invincible! Always attack. Security 
is the greatest enemy. Always be alert. Ex- 
ecute silently, 8 

Not to be outdone, General Kanda had 
this to say: 

We must fight to the end to avenge the 
shame of our country's humiliation on 
GUADALCANAL. . . . There can be no 
rest until our bastard foes are battered, and 
bowed in shame— till their . . . blood adds 
lustre ... to the badge of the Sixth Di- 
vision. Our battle cry will be heard 
afar. . . . 9 

Again, the most apt comment is in 
Proverbs XVI: 18. 

Hill joo 

At 0545, shortly after daybreak of 8 
March, Hyakutake's artillery heralded 
the opening of his counteroffensive by 
firing on all parts of the beachhead, with 
especial attention to the Piva airfields. 

American observers on the ground, in 
artillery liaison planes, and on board de- 
stroyers, aided by information gained 
from documents, quickly determined the 
general location of the Japanese artillery, 
and counterbattery fire by the corps artil- 
lery and the organic division artillery 
battalions began at once. The Americal 
Division artillery put its fire on hills to 
the east and east-northeast, the 37th on 

8 Quoted in Frankel, The jyth Infantry Division 
in World War II, pp. 142—43. The "red emblem" 
referred to was probably the shoulder patch of. the 
6th Division. 

" Quoted in Griswold, Bougainville, p. 81. 

those to the northeast. Smoke shells were 
fired at suspected enemy observation 
posts to blind the enemy. In the 37th's 
sector the 6th Field Artillery Battalion, 
supporting the 129th Infantry, and the 
129th Infantry Cannon Company were 
so situated that they could shoot directly 
at enemy gun flashes. The other battal- 
ions fired by forward observer. 10 

At 1045 twenty-four SBD's and twelve 
TBF's of the 1st Marine Air Wing 
dropped fourteen tons of bombs on Hills 
250 and. 600. A strike against Hill 1111 
was planned for the late morning but 
was postponed when a sudden cloud 
screen obscured the hilltop. Finally, at 
1600, fifty-six SBD's and thirty-six TBF's, 
guided by artillery smoke shells, dropped 
100- and 1,000-pound bombs on Hill 
1111 and environs. 

In the course of the day's firing the 
Japanese destroyed one B-24 and three 
fighters, and damaged nineteen planes 
on Piva strips. Before nightfall all 
bombers except six TBF's which re- 
mained for local support left for New 
Georgia to escape destruction. The 
enemy also damaged one 155-mm. gun 
and several tanks. Early next morning, 
the 9th, the enemy guns turned their 
attention to the Torokina fighter strip 
and forced its planes to take to the air 
for safety. Almost no shells fell on the 
front lines except in the 145th In- 
fantry's area, where shellfire and mor- 
tars caused several casualties. 

The sector of the 145th, now com- 
manded by Col. Cecil B. Whitcomb, ex- 
tended from low ground in the vicinity 
of the Numa Numa Trail eastward past 

"Throughout the operation U.S. destroyers also 
fired counterbattery fire and against suspected 
enemy assembly areas and approach routes. 



the south shore of Lake Kathleen and up 
along the military crest of Hill 700, a 
frontage of about 3,500 yards. The 3d 
Battalion, on the left (west), held the 
low ground just south of Lake Kathleen 
and Cannon Hill, an eminence slightly 
lower and to the west of Hill 700. On 
the right the 2d Battalion held Hill 700 
with two rifle companies (E and G) and 
machine gun sections of H Company in 
line, F Company in reserve, and H Com- 
pany's 81-mm. mortars grouped on the 
reverse (south) slope. 

Hill 700, which commanded the en- 
tire beachhead, was steep, with slopes of 
65 to 75 percent in all directions. Amer- 
ican intelligence estimates, though not 
ruling out an enemy attack here, had 
tended to discount its probability. The 
steepness that increased the difficulty of 
attack also complicated the defense, for 
the forward (north) slope fell away too 
sharply to permit it to be completely 
covered with grazing fire. Thus the 2d 
Battalion had an extra allotment of ma- 
chine guns. Its pillboxes housed 37-mm. 
antitank guns, light and heavy machine 
guns, BAR's, and rifles. The front was 
wired in, with some mines in front. In 
direct support were the 105-mm. how- 
itzers of the 135th Field Artillery Bat- 
talion and, starting on 8 March, the 4.2- 
inch mortars of D Company, 82d Chem- 
ical Battalion. 11 

That the 145th Infantry was in danger 
of attack had become obvious on 6 March 
when patrols reported the presence of 
large numbers of Japanese about four- 
teen hundred yards north of Hill 700. 
Additional ammunition was made avail- 
able to the troops, and two days' C 

" This company also supported Che 129th In- 

rations, ammunition, and a five-gallon 
can of water were stocked in each pill- 
box against a breakdown in supply. For 
nocturnal illumination each machine 
gun section was issued four incendiary 
grenades and a gallon can of flame 
thrower fuel. 

On 7 March Japanese wire-cutting 
parties started work in front of the 145th. 
Next day patrols in front of Hill 700, 
Cannon Hill, and along Lake Kathleen's 
shores kept running into enemy troops. 
At the same time 129th Infantry patrols 
reported many enemy contacts, and 
Americal Division patrols also observed 
enemy troops east of the Torokina River, 
along the East-West Trail, and around 
Hills 250 and 600. In front of the 145th 
fire fights and skirmishes went on all day. 
When patrols reported that the enemy 
was massing, the 37th Division artillery, 
the i45th's Cannon Company, and the 
4.2-inch mortars fired a counterprepara- 
tion twelve hundred yards wide and two 
thousand yards deep in front of the 2d 
Battalion. Japanese orders had called for 
an attack on 8 March, but none de- 
veloped. The 23 d Infantry had spent the 
day moving into position in front of the 
145th; the 2d Battalion reconnoitered 
Cannon Hill, the 3d, 700, but for some 
reason the regiment did not assault. 12 

Rain fell throughout the night of 8-g 
March. Shortly after midnight, concealed 
by darkness, rain, and mists, about two 
companies of the 23d Infantry attacked 
up the north slope of Hill 700 against 
the 1st Platoon, G Company, 145th, 

" In 1949 General Kancla said that as the 2jd's 
attack had not succeeded, and as the 13th had taken 
Hill 260 on 8 March, he ordered the attack con- 
tinued on the 9th. Either his recollection was faulty 
or his subordinate commanders deceived him. 

37 th Division Troops Carrying 5-Gallon Cans of Water up the steep slope 
of Hill joo. Foreground, from left: Pfc. Howard K. Stoneburner, Sgt. Brant A. 
Johnson, and Pfc. Thomas J. Householder. Background, from left: Pfc. Gerald 
C. Menken, Pfc. Taylor Maggard, and S/Sgt. Robert G. Dove. 



which held a level saddle between the 
topmost eminence of the hill and a rise 
to the left (west) dubbed Pat's Nose. 
Other elements of the 23 d put pressure 
on E Company, 145th, on the highest 
point of 700. This attack was repulsed. 

About 0230, 9 March, the 23d In- 
fantry attacked G Company's 1st Pla- 
toon again, this time in column of bat- 
talions. The 2d Battalion, in the lead, 
blew up the barbed wire, knocked out a 
pillbox, and through the gap its forward 
elements moved onto the saddle and set 
up machine guns. American mortars and 
artillery opened up and appear to have 
severely punished the 3d Battalion, 
which was following the 2d. 

When day broke the Americans were 
not sure of the extent of the enemy 
penetration, as mists and enemy fire 
hampered reconnaissance. Some local 
counterattacks, largely un-co-ordinated, 
were attempted but all failed. Soldiers 
of the 145th tried to attack northward 
up the south slopes of Hill 700, but the 
Japanese drove them back by rolling 
down grenades. 

By noon the situation was clarified. 
The Japanese had made but a minor 
penetration; about one company held a 
salient on the saddle about one hundred 
yards from east to west and fifty yards 
deep. It had captured seven pillboxes, 
plus observation posts, in the 1st Pla- 
toon's line and had set up light and 
heavy machine guns. 

General Beightler released the 1st Bat- 
talion, 145th, from division reserve to 
Colonel Whitcomb, who attached it to 
the 2d Battalion, and elements of the 
117th Engineer Battalion took up posi- 
tions in the i45th's regimental reserve 
lines south of Hill 700. 

About noon C Company, 145 th, 
started northward up Hill 700 toward 
the saddle in frontal assault while two F 
Company platoons attacked the saddle 
from the east and west. By 1530 the pla- 
toon attacking from the east had re- 
covered some of the lost ground but C 
Company had been halted about two 
thirds of the way to its objective. 

Two light tanks of the 754th Tank 
Battalion, released out of corps reserve 
by General Griswold, tried to support 
an attack later in the afternoon, but the 
hill proved too steep for them. The F 
Company platoons pressed their attack 
anyway and by 1735 had retaken five pill- 
boxes. By nightfall a solid line had been 
established in front of the Japanese. It 
ran along Hill 700 south of the crest in 
the region of the penetration and joined 
its flanks with the original main line of 
resistance. B and C Companies and one 
platoon of D Company held the new 

During the day the Japanese used 
their point of vantage on the saddle to 
put mortar and machine gun fire on Mc- 
Clelland Road, a lateral supply route 
south of the crests of the hills, roughly 
parallel to the main line of resistance. 
This fire halted the %-ton trucks and 
half-tracks that were used to bring up 
ammunition and required the use of 
hand-carrying parties, which hauled am- 
munition forward and took out the 
wounded under Japanese fire. 

Neither Japanese nor Americans made 
any aggressive moves on the night of 
g-10 March, but it was a noisy night. 
The Japanese laid mortar and small arms 
fire on the American lines, while the 37th 
Division artillery and mortars put close- 



in and deep supporting fires in front of 
the 2d Battalion, 145th. 

At 0645 the next day, 10 March, while 
the Muda Unit began its attack against 
the Americal Division troops on Hill 
260, the 23d Infantry troops on the sad- 
dle renewed their attack and other ele- 
ments of the Iwasa Unit attempted to 
get through the curtain of American 
artillery and mortar fire to reinforce the 
saddle. The Americans on Hill 700 re- 
sponded with fire and local counter- 
attacks. There was no change in the loca- 
tion of the front lines. 

During the morning Griswold re- 
leased the Provisional Infantry Battalion, 
251st Antiaircraft Artillery Regiment, 
from corps reserve; it proceeded to the 
i45th's regimental reserve line. Elements 
of the 117th Engineer Battalion there- 
upon made ready to destroy the Japanese 
positions with bangalore torpedoes and 
pole charges of TNT. But this came to 
naught when four engineers, trying to 
snake a torpedo into a pillbox, were 
killed outright by the torpedo, which 
either exploded prematurely or was 
detonated by a Japanese shell. A Japa- 
nese-speaking American soldier brought 
a loud-speaker up close to the enemy and 
urged immediate capitulation; the Jap- 
anese responded with a mortar shell 
which knocked the loud-speaker out of 

By afternoon of this day of patternless 
and ineffective action (which also fea- 
tured enemy fire on McClelland Road 
and a 36-plane strike against Japanese 
positions), the American units in contact 
with the enemy had become inter- 
mingled. Sorting and reorganizing them 
consumed much of the afternoon, so that 
it was 1700 before elements of the 1st 

and 2d Battalions, 145th, after a ten- 
minute mortar preparation, delivered a 
co-ordinated attack. The Americans used 
bangalore torpedoes, rocket launchers, 
and pole charges in the face of artillery, 
mortar, machine gun, and rifle fire. The 
fighting was close work; several pillboxes 
were recaptured and then lost. As dark- 
ness fell, however, the Americans had 
achieved some success. The Japanese 
penetration was now reduced by more 
than half. By 1930, G, F, A, C, B, and E 
Companies held the line; the 37th Re- 
connaissance Troop, which General 
Beightler attached to the 145th at 1815, 
was in reserve. 

During the night, as Colonel Magata 
prepared to deliver his attack against 
the 129th Infantry, General Iwasa sent 
the rest of his command against the 
i45th's front from Cannon Hill to the 
crest of 700. The Japanese came in 
closely packed waves, shouting, the 37th 
Division reported, imprecations in Japa- 
nese. The fields of fire at Cannon Hill 
and Pat's Nose were better than at 700, 
and the 145th, heavily supported by 
artillery and mortars, handily repulsed 
Iwasa everywhere except on the saddle, 
where the Japanese captured one more 

As dawn broke on 1 1 March, a day on 
which Muda was active at Hill 260 and 
the Magata Unit began its attack against 
the 129th Infantry, General Beightler 
was obviously concerned over the i45th's 
failure to reduce the enemy salient. The 
night before he had ordered the 2d Bat- 
talion, 148th Infantry, to move from its 
regimental reserve positions to the 
i45th's sector. To replace it General 
Griswold placed the 1st Battalion, 24th 
Infantry, at Beightler's disposal. Beight- 



Two Light Tanks M3 of the 754™ Tank Battalion heading up Hill yoo 
during the afternoon of 9 March. 

ler also dispatched his assistant division 
commander, Brig. Gen. Charles F. Craig, 
to the i45th's sector to observe opera- 
tions and keep him informed. The regi- 
mental commander was suffering, Craig 
reported later, from extreme battle 
fatigue and was relieved. Colonel Freer, 
who had been serving as executive o£ 
the 145th, took his place. 13 

In the meantime the Japanese made 
valorous efforts to put more troops onto 
the saddle. The Americans resisted with 
vigor and with all the fire power at their 
disposal. Charging, literally, over the 
piled heaps of their dead comrades, the 

"Ltr, Gen Craig to Gen Smith, Chief of Mil 
Hist, go Oct 53, no sub, OCMH. 

enemy soldiers fought hard but vainly, 
and failed either to budge the Americans 
or to strengthen the saddle. 

The 2d Battalion, 148th, reached its 
assembly area behind the 145th at 1115. 
Colonel Radcliffe, its commander, re- 
connoitered in preparation for an after- 
noon attack. 

Three 105 -mm. howitzer battalions, 
the 145th Infantry Cannon Company, 
4.2-inch chemical mortars, the 81-mm. 
mortars of D, H, and M Companies of 
the 145 th, and the 60-mm. mortars of all 
the rifle companies of the 2d Battalion, 
148th, fired a preparation from 1320 to 
1330. Then elements of the 148th at- 
tacked. Two platoons from E Company 



moved east from Pat's Nose in an effort 
to envelop the saddle from north and 
south while a third platoon delivered a 
holding attack, westward from the crest 
of Hill 700. The whole target was 
blanketed by artillery smoke shells. The 
145th supported the attack with over- 
head fire. The platoon making the en- 
velopment from the north gained the 
crest, losing eight dead, whereupon the 
platoon leader and four enlisted men 
seized a communication trench, then a 
pillbox. But the Japanese killed the five 
men and the attack halted about 1900. 
The troops dug in on the ground they 
had gained. During the night the Japa- 
nese harried the Americans but failed to 
penetrate the line. 

The attack on 1 2 March followed the 
previous day's pattern. While intense 
local battles raged in the 129th Infantry 
sector and on Hill 260, E Company con- 
tinued its attack and F Company at- 
tacked northwestward from the top of 
700. Using grenades, rifles, flame throw- 
ers, and rocket launchers, the 148th 
soldiers methodically reduced the pill- 
boxes one by one. When nearly all the 
officers in both companies were wound- 
ed, sergeants took over command. By 
1300 the Japanese held but one pillbox; 
by 1317 they had lost it, and by 1530 
mopping up was completed, all the Jap- 
anese save two wounded prisoners were 
dead, the i45th's line was restored. Three 
hundred and nine enemy corpses were 
counted in the immediate area. During 
the next day the Iwasa Unit, which had 
suffered heavily in its unsuccessful at- 
tack, withdrew behind a screen of com- 
bat patrols and fire. 

During the period 8-13 March the 
37th Division lost five officers and sev- 

enty-three enlisted men killed. 14 The ar- 
tillery expended a considerable amount 
of ammunition in defense of Hill 700: 
20,802 105-mm. rounds; about 10,000 75- 
mm. rounds; 13,000 81-mm. and 811 4.2- 
inch mortar shells. 15 

Hill 260 

While General Iwasa was meeting de- 
feat at Hill 700, Colonel Muda was at- 
tacking the American outpost on Hill 
260 in preparation for operations against 
Hills 309 and 608 in the Americal Divi- 
sion's sector. The Muda Unit— princi- 
pally one battalion and two companies 
of the 13th Infantry— completed its as- 
sembly at Peko and moved forward. On 
the night of 9-10 March small enemy 
forces infiltrated between Hill 260 and 
the main line of resistance, while an as- 
sault force assembled east of 260 and 
made ready to attack. 16 

Some 800 yards east of the main perim- 
eter line and 7,500 yards north of the 
Torokina's mouth, Hill 260 is shaped 
like an hourglass. Its long axis runs from 
northwest to southeast. The two ends 
of the hourglass are rises called North 
Knob and South Knob. Each knob is 
about half the size of a football field. 
The handle between them is slightly 

" Memo, G-i 37th Div for G-a and G-3, 14 Mar 
44, no sub, in 37th Div G-3 Jnl File, Vol. 13, Serials 

"On 14 March* General Kreber ordered 90-mm. 
antiaircraft guns to supplement certain 40-mm. 
guns already in use on the front lines. Thereafter 
these flat-trajectory weapons sniped at enemy guns 
emplaced on the forward slopes of the hills to the 
north and northeast. 

18 In addition to sources cited above, this account 
of the Hill 260 action is based on comments of Col. 
William D. Long on draft MS of this volume, at- 
tached to his Ltr to Gen Smith, Chief of Mil Hist, 
21 Mar 54, no sub, OCMH. 



lower and so narrow that there was room 
for only a trail. North and South Knobs 
lie so close together— less than 150 yards 
apart— that to hit one knob with artillery 
or mortar fire inevitably showered the 
other with fragments. The slopes to the 
east and west are very steep. 

The East- West Trail crossed the Tor- 
okina just east of 260 and, bending south 
of South Knob, entered the main perim- 
eter line between Hills 3og and 608. A 
small north-south stream, called the 
Eagle River by the Americans, flowed 
between 260 and the main perimeter 
before running into the Torokina River. 
In the early part of March only one trail 
led from the main line to South Knob. 
The last hundred yards to the top con- 
sisted of a steep stairway revetted into 
the southwest slope. A small vehicular 
bridge had been built over the Eagle. 
The entire area, including the east and 
west slopes of 260, was heavily jungled. 

From a platform on a 150-foot-high 
tree ("OP Tree") on South Knob, Amer- 
ican mortar and artillery forward ob- 
servers could view the banks of the Tor- 
okina below, the East-West Trail, and 
Hills 250 and 600 to the northeast. Con- 
versely, in Japanese hands Hill 260 
would have provided good observation 
of Hills 608 and 309 and of the corps' 
rear area between them. On 10 March 
the American garrison on 260 consisted 
of about eighty men including forward 
observers and a reinforced platoon from 
G Company, i82d Infantry, which regi- 
ment held the main perimeter west of 
260. It was "a sore thumb stuck out into 
the poison ivy." 17 Defenses, all on 
South Knob, consisted of pillboxes and 

" Ibid. 

"OP Tree" on Hill 260 

bunkers inside barbed wire and defen- 
sive warning devices arranged in a tri- 
angle around OP Tree. Fire lanes faced 
northeast, east, and southeast. 

The Americans had been maintaining 
ambushes on the northeast and south- 
east approaches to South Knob but on 
9 March withdrew them to put harassing 
and interdictory fires over the whole 
area from 1830 to 2330, 9 March, and 
again from 0415 to 0500, 10 March. A 
few minutes after 0600 on 10 March, 
during the i82d's normal stand to for 
the two hours before daylight, fire from 
Japanese mortars, machine guns, and 



rifles began striking the American posi- 
tions on 260. At 0638 an officer of the 
246th Field Artillery Battalion, Ameri- 
cal Division, reported from his post in 
OP Tree that the Japanese had attacked 
and were all around the base of his tree. 
He was not heard from again. The at- 
tack, which the Americans estimated was 
made by one company, was actually de- 
livered by all or part of the 3d Battalion , 
13th Infantry. It overran most of the 
American positions, captured OP Tree, 
and drove the survivors of the American 
garrison to North Knob. One six-man 
group from the mortar and artillery ob- 
servation teams took refuge in two pill- 
boxes and put up such stout-hearted re- 
sistance that they held their positions in 
spite of the fact that the Japanese had 
surrounded them. 18 

When the enemy attack was reported 
to General Griswold, he ordered General 
Hodge, the Americal Division's com- 
mander, to hold 360 at all costs. This 
order came as a surprise to the Americal's 
officers, who had not expected to be re- 
quired to hold 260 in the face of a strong 
enemy attack. 1 " Col. William D. Long, 
commanding the i82d, promptly re- 
leased two companies— E and F— of his 
?d Battalion from regimental reserve 
and placed Lt. Col. Dexter Lowry, com- 
mander of the 2d Battalion, in charge of 

18 For a stirring, well-told account by one of the 
observers, see Griswold, Bougainville, pp. 103-09, or 
SOPACBACOM, The Bougainville Campaign, Ch. 
VIII, pp. 368-75, OCMH. Colonel Long, in his 
comments on the draft MS of this volume, expressed 
his admiration for the skill of the Japanese who 
delivered the attack "from the most difficult direc- 
tion with complete surprise." 

18 Comments on draft MS of this volume by Lt 
Col Carl D. McFerren, formerly ACofS, G-2, 
Americal Div, on Bougainville, in OCMH. 

operations. 20 F Company left the perim- 
eter, crossed the Eagle River and pushed 
northward through virgin jungle to 
North Knob, made contact with the G 
Company soldiers who had made their 
way there, and established a perimeter 
defense. At 0845 E Company was ordered 
to advance east over the trail to attack 
the South Knob from the southwest in 
conjunction with a southward move by 
one platoon of F Company. 

By 1045, when E Company reached 
the base of Hill 260's southwest slope, 
the troops on North Knob had become 
aware that some Americans on South 
Knob were still alive. 21 The attack be- 
gan immediately after E Company's ar- 
rival. One E Company platoon started 
up the steep slope as the F Company 
platoon attempted to move south, but 
after a gain of about thirty-five yards 
both platoons, now coming into the 
cleared areas, halted under enemy fire. 
Shortfy after 1300 Long authorized 
Lowry to contain the Japanese at the 
base of OP Tree until he could send 
flame throwers forward. 

Lowry therefore held up the attack, 
received the flame throwers, and by 1420 
was ready to go again. This time he 
planned a double envelopment from 
260's southwest spur. One platoon of E 
Company was to move left (north) to 
make contact with the F Company pla- 
toon advancing south while a second 
E Company platoon moved right to at- 
tack from the south and southeast. The 
platoons moved out and began their at- 

n Long had served as ACofS, G— 2, Americal Di- 
vision, during the Guadalcanal Campaign. 

" They apparently based this conclusion on fire 
fights between the Japanese and the trapped ob- 



tack at 1445. The Japanese quickly 
halted the northern attack. The south- 
ern platoon started up South Knob, met 
grazing fire, retired, moved to the right, 
and assaulted again. This time, using 
flame throwers and grenades, the pla- 
toon drove up onto a shelf on the south- 
ern edge of South Knob that protected 
it from small arms fire. It was within 
earshot of the trapped Americans. Col- 
onel Lowry, now estimating that at least 
two enemy companies held South Knob, 
reported that his and the enemy's forces 
were too close for him to use 60-mm. 
mortars safely. The attack was renewed 
at 1800, but by then battle casualties 
and exhaustion had reduced E Com- 
pany's strength by one half. Lowry and 
Colonel Long, who had arrived at South 
Knob at 1715, decided to hold their 
present positions. The six Americans 
in the pillboxes thought E Company had 
secured the hill, and stayed where they 
were. Active operations for the day were 
concluded by an enemy bayonet assault 
which F Company repulsed by fire. 

Early next morning the Japanese, ap- 
parently strengthened during the night, 
struck at E Company in a quick attack. 
The company turned back the attackers, 
but of its 7 officers and 143 enlisted men 
who had left the perimeter the day be- 
fore, only 1 wounded officer and 24 en- 
listed men remained at the front. Col- 
onel Long therefore ordered G Com- 
pany (less its original outpost platoon) 
out of the main perimeter to relieve E. 

As G Company advanced up the west 
slope of South Knob it ran into enemy 
troops attacking from the east, south- 
east, and south. Colonel Lowry reported 
his troops in distress as the enemy threat- 
ened to encircle his position on South 

Knob and began driving E Company off. 
But there was almost no way to 
strengthen him. General Hodge was re- 
quired to hold one battalion available